Phytochemical and Biological Investigation of
Immature Conifer Cones for
Antibacterial and Modulatory Activity against
Multidrug-resistant and Methicillin-resistant
Staphylococcus aureus
Thesis presented by
Eileen Catherine Jean Smith
for the degree of
Doctor of Philosophy
Centre for Pharmacognosy and Phytotherapy
The School of Pharmacy
University of London
2006 ProQuest Number: 10105105
All rights reserved
INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest.
ProQuest 10105105
Published by ProQuest LLC(2016). Copyright of the Dissertation is held by the Author.
All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code. Microform Edition © ProQuest LLC.
ProQuest LLC 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 This thesis describes research conducted in the School of Pharmacy, University of London between April 2002 and January 2006 under the supervision of Dr. Simon Gibbons and Professor Elizabeth Williamson. I certify that the research described is original and that any parts of the work that have been conducted by collaboration are clearly indicated. I also certify that 1 have written all the text herein and have clearly indicated by suitable citation any part of this dissertation that has already appeared in publication.
,.lc^/\ 9 /i Signature Date
- 2 - ABSTRACT
Antibiotic resistance by pathogenic bacteria is a major problem both in hospitals and in the community. Of particular concern is methicillin-resistant Staphylococcus aureus (MRSA), many strains of which have acquired resistance to most antibiotics. Another mode of resistance is by means of an efflux pump and many S. aureus strains have acquired pumps which confer multidrug-resistance by effluxing many different compounds out of the cell. There is an urgent need to find new antibacterials and new ways to fight these resistant strains.
The rationale for this study is that since cones are essential for reproduction in conifers, the immature cones are likely to contain compounds which protect against microbial invasion. Initial screening of cones from several conifer species identified anti-staphylococcal activity, which was greatest in the hexane extracts.
Bioassay guided fractionation and structure elucidation using 1-D and 2-D NMR yielded several active diterpenes from Chamaecyparis lawsoniana, Chamaecyparis nootkatensis and Pinus nigra. These compounds showed activity (2 - 64 pg/ml) against multidrug-resistant and effluxing S. aureus clinical isolates, and against epidemic MRSA strains EMRSA-15 and -16, which are the major strains found in UK hospital MRSA bacteraemias.
Some of the isolated diterpenes also demonstrated activity as potentiators of antibiotic activity. Ferruginol restored oxacillin sensitivity in EMRSA-15, and moderate activity in potentiating antibiotic activity against effluxing strains was also observed for ferruginol and totarol. Efflux inhibition assays suggested that these compounds were weak efflux pump inhibitors.
This study demonstrates that compounds from immature conifer cones have good antibacterial activity and some modulatory activity against resistant strains of S. aureus. These compounds are worthy of further investigation, particularly as plants produce compounds which clinically relevant bacteria are unlikely to have previously encountered.
3 - TABLE OF CONTENTS Page ABSTRACT 3
TABLE OF CONTENTS 4
TABLE OF FIGURES 9
TABLE OF TABLES 11
ABBREVIATIONS 13
ACKNOWLEDGEMENTS 14
1.0 INTRODUCTION 15
1.1 THE PROBLEM OF RESISTANT BACTERIA 15 1.2 ANTIBACTERIALS 16 1.2.1 Triclosan 17 1.2.2 Silver Ions 19 1.3 ANTIBIOTICS 20 1.3.1 P-lactams 20 1.3.2 Glycopeptides 22 1.3.3 Macrolides 23 1.3.4 Tetracyclines 24 1.3.5 Fluoroquinolones 25 1.4 METHODS OF BACTERIAL RESISTANCE 26 1.4.1 Intrinsic Resistance 26 1.4.2 Acquired Resistance 27 1.4.3 Efflux Pumps 28 1.4.4 iS. Efflux Pumps 30 1.5 STAPHYLOCOCCI 32 1.5.1 Methicillin-resistant *S. (MRSA) 33 1.5.2. Penicillin Binding Protein (PBP2') 37
1.6 . NEW ANTIBIOTICS 39 1.7 MODULATION 41
- 4 - 1.8 CONIFERS 43 1.8.1 Chamaecyparis 43 1.8.2 Chamaecyparis lawsoniana 44 1.8.3 Chamaecyparis nootkatensis 44
1.8.4 'EûmohoXdûCvy0 Ï Chamaecyparis 46 1.8.5 Ch.Qm\s\iy o f Chamaecyparis 47 1.8.6 Pinus 51 1.8.7 Pinus nigra 51 1.8.8 Ethnobotany of Pines 52 1.8.9 Chemistry of Pines 52 1.8.10 Present Day Use of Conifer Compounds 55 1.8.11 Conifer Oleoresin 57 1.9 TERPENES 59
2.0 MATERIALS AND METHODS 2.1 PHYTOCHEMICAL METHODS 63 2.1.1 Conifer Material 63 2.1.2 Solvent Extraction 63 2.1.3 Thin-layer Chromatography (TEC) 64 2.1.4 Preparative Thin Layer Chromatography (PTLC) 65
2.1.5 Vacuum Liquid Chromatography (VLC) 6 6 2.1.6 Biotage"^^ Flash Chromatography 67 2.1.7 Solid Phase Extraction (SPE) 67
2.1.8 Size Exclusion Chromatography 6 8 2.1.9 High Performance Liquid Chromatography (HPLC) 69 2.2 SPECTROSCOPIC METHODS 70 2.2.1 Nuclear Magnetic Resonance (NMR) 70 2.2.1.1'HNMR 71 2.2.1.2'^CNM R 72 2.2.1.3 Two-dimensional spectra 72 2.2.1.4 Correlation Spectroscopy (COSY) 72 2.2.1.5 Heteronuclear Single Quantum Coherence (HSQC) 73 Spectroscopy
2.2.1.6 Heteronuclear Multibond Coherence (HMBC) 73 Spectroscopy
5 - 2.2.1.7 Nuclear Overhauser Effect Spectroscopy (NOESY) 73 2.2.2 Infra-red Spectroscopy and Polarimetry 74 2.2.3 Gas Chromatography-Mass Spectrometry (GC-MS) 74 2.3 BIOLOGICAL METHODS 75 2.3.1 Minimum Inhibitory Concentration (MIC) Assay 75 2.3.2 Modulation Assay 76 2.3.3 S. aureus Strains used in MIC and Modulation Assays 77 2.3.4 Ethidium Efflux Assay 78
3.0 RESULTS 79 3.1 PRELIMINARY SCREENING 79 3.2 EXTRACTION OF COMPOUNDS FROM CHAMAECYPARIS 82 LAWSONIANA 3.3 PHYTOCHEMICAL CHARACTERISATION OF COMPOUNDS 8 6 FROM CHAMAECYPARIS LAWSONIANA 3.3.1 CL-001 (Ferruginol) 86 3.3.2 CL-002 (Pisiferol) and CL-003 (5-Epipisiferol) 93 3.3.3 CL-004 (Formosanoxide) 101 3.3.4 CL-005 (4p-Hydroxygermacra-l(10),5-diene) 107
3.3.5 CL-006 (^raw5-Communic acid) 111 3.3.6 CL-007 (Torulosal) 118 3.3.7 CL-008 (Oplopanonyl acetate) 123 3.4 EXTRACTION OF COMPOUNDS FROM CHAMAECYPARIS 127 NOOTKATENSIS 3.4 PHYTOCHEMICAL CHARACTERISATION OF COMPOUNDS 130 FROM CHAMAECYPARIS NOOTKA TENSIS 3.5.1 CN-001 (Totarol) 130 3.5.2 CN-002 (7a-Hydroxytotarol) 135 3.5.3 CN-003 (Sempervirol) 141 3.5.4 CN-004 (Sugiol) 145 3.6 PINUSNIGRA 150 3.6.1 Preliminary Screening of Six Pinus Species 150 3.6.2 Extraction of PN-001 from «/gra 152 3.6.3 PN-001 (Isopimaric acid) 153
6 - 4.0 MIC AND MODULATION ASSAYS - RESULTS 159 AND DISCUSSION 4.1 RESULTS OF ANTI-STAPHYLOCOCCAL MIC ASSAYS 159 4.2 RESULTS OF MODULATION ASSAYS 165 4.3 ASSAYS FOR POTENTIATION OF COMPOUND ACTIVITY 170 4.3.1 Results 170 4.3.2 Abietic Acid and Reserpine 171 4.3.3 Molecular Modelling 173 4.4 RESULTS OF EFFLUX INHIBITION ASSAYS 176 4.4.1 The Effect of Ferruginol on Efflux of Ethidium Bromide 176 4.4.2 The Effect of Totarol on Efflux of Ethidium Bromide 177 4.5 DISCUSSION 179 4.5.1 Ferruginol 179 4.5.2 Totarol 182 4.5.3 Sempervirol 187 4.5.4 Sugiol 187 4.5.5 Pisiferol, 5-Epipisiferol and Formosanoxide 188
4.5.6 /rfl«5 -Communic Acid 189 4.5.7 4(3-Hydroxygermacra-l(10),5-diene 190 4.5.8 Isopimaric Acid and Abietic Acid 191 4.5.9 Oplopanonyl Acetate and Torulosal 192 4.6 STRUCTURE ACTIVITY RELATIONSHIPS 193 4.6.1 Phenol and Isopropyl Groups 194 4.6.2 Lipophilicity and SARs 195 4.6.3 Oxidation State of Functional Groups 197 4.6.4 Pisiferol Epimers and Formosanoxide 199 4.6.5 Potentiation of Antibiotic Activity - SARs 201
5.0 PEDICULICIDE ASSAYS AGAINST HUMAN LICE 208 5.1 INTRODUCTION 208 5.1.1 Head Lice Treatments 210 5.1.2 The Orlando Strain of Clothing Lice 213 5.2 METHODS 215 5.2.1 Filter Paper Disc Assay 215 5.2.2 Dip Method Pediculicide Assay 216
- 7 - 5.3 RESULTS 218 5.4 DISCUSSION 225
6.0 CONCLUSION 230
REFERENCES 232
PUBLICATIONS 267
- 8 - TABLE OF FIGURES Page Figure 1.2.2: Treatment of an MRSA infected leg ulcer with silver dressing 19 Figure 1.3.1: P-lactam ring with a fused thiazolidine ring 2 1 Figure 1.3.3: Erythromycin 23 Figure 1.3.4: Tetracycline 24 Figure 1.3.5: Norfloxacin 25 Figure 1.4.3: Drug efflux in Gram-positive and Gram-negative bacteria 30 Figure 1.5a: S. aureus 32 Figure 1.5b: S. aureus biofilm 32 Figure 1.5.1a: Methicillin 33 Figure 1.51.b: Headlines from selected articles on MRSA (2004/05) from 36 the London Evening Standard Figure 1.5.2: The precursors required for PBP2' activity 38 Figure 1.6: Linezolid 39 Figure 1.8.2a: C. lawsoniana 44 Figure 1.8.2b: C. lawsoniana shoot 44 Figure 1.8.3: C. nootkatensis shoot 45 Figure 1.8.5a: terpinen-4-ol and sabinene 43 Figure 1.8.5b: (+)-5-cadinene, (-)-p-curcumene and nootkatene 48 Figure 1.8.5c: Ferruginol, pisiferic acid and pisiferdiol 48 Figure 1.8.5d: Nootkastatin 2 and quercetin 49 Figure 1.8.5e: Chamaecydin 49 Figure 1.8.5f: Chamalignolide 50 Figure 1.8.6a: P. nigra 51 Figure 1.8.6b: Immature cone and shoot of P. nigra 51 Figure 1.8.9a: a and p-pinene and limonene 53 Figure 1.8.9b: Examples of abietane, pimarane and labdane resin acids 54 Figure 1.8.9c: Pinosylvin 54 Figure 1.9a: Isopentenyl pyrophosphate and dimethylallyl pyrophosphate 59 Figure 1.9b: Biosynthesis of GPP 60 Figure 1.9c: Formation of the sesquiterpene precursor FPP 61 Figure 1.9d: Geranyl geranyl pyrophosphate 61 Figure 3.2a: SPE of VLC fraction 4 83 Figure 3.2b: TLC of pooled sephadex column fractions 84 Figure 3.2c: Bioassay guided fractionation of C. lawsoniana 85 Figure 3.3.1a: CL-001 (ferruginol) 8 6 Figure 3.3.1b: ^^C carbon and DEPT-135 spectra of CL-001 87 Figure 3.3.1c: ^H-NMR spectra of CL-001 8 8 Figure 3.3.Id: NOESY spectrum for CL-001 91 Figure 3.3.1e: COSY and HMBC spectra of CL-001 92 Figure 3.3.2a: CL-002 (pisiferol) 93 Figure 3.3.2b: CL-003 (5-epipisiferol) 93 Figure 3.3.2c: Selected NOEs for CL-002 and CL-003 97 Figure 3.3.2d: NOESY spectra for CL-002 (top) and CL-003 99 Figure 3.3.2e: ^H spectra for CL-002 (top) and CL-003 1 0 0 Figure 3.3.3a: CL-004 (formosanoxide) 1 0 1 Figure 3.3.3b: Proton spectrum of CL-004 104 Figure 3.3.3c: NOESY spectrum of CL-004 105
- 9 - Figure 3.3.3.d: GC-MS data for CL-004 106 Figure 3.3.4a: CL-005 (4p-hydroxygermacra-l(10),5-diene) 107 Figure 3.3.4b: spectra of CL-005 110 Figure 3.3.5a: CL-006 (^ra«5 -communic acid) 111 Figure 3.3.5b: spectrum of CL-006 117 Figure 3.3.6a: CL-007 (torulosal) 118 Figure 3.3.6b: spectrum of CL-007 120 Figure 3.3.6c: NOESY spectrum of CL-007 122 Figure 3.3.7a: CL-008 (oplopanonyl acetate) 123 Figure 3.3.7b: spectrum of CL-008 126 Figure 3.4.1 : HPLC trace of CHCI3 VLC fraction 3 129 Figure 3.5.1a: CN-001 (totarol) 130 Figure 3.5.1b: spectrum for CN-001 133 Figure 3.5.1c: NOESY spectrum of CN-001 134 Figure 3.5.2a: CN-002 (7a-hydroxytotarol) 135 Figure 3.5.2b: GC trace of CN-001 136 Figure 3.5.2c: GC trace of CN-002 136 Figure 3.5.2d: Mass spectrum of CN-001 137 Figure 3.5.2e: Mass spectrum of CN-002 137 Figure 3.5.2f: NOESY spectrum of CN-002 139 Figure 3.5.2g: NOE between H-7 and H-15 of CN-002 140 Figure 3.5.2h: *H spectrum of CN-002 140 Figure 3.5.3a: CN-003 (sempervirol) 141 Figure 3.5.3b: spectrum of CN-003 144 Figure 3.5.4a: CN-004 (sugiol) 145 Figure 3.5.4b: ’H spectrum of CN-004 148 Figure 3.6.1: TLC of Pinus species hexane extracts 150 Figure 3.6.2: TLC of pooled flash chromatography fractions 152 Figure 3.6.3a: PN-001 153 Figure 3.6.3b: spectrum for PN-001 157 Figure 3.6.3c: NOESY spectrum of PN-001 158 Figure 4.3.2a: Structure of abietic acid and reserpine 173 Figure 4.3.3a: Molecular model of abietic acid-reserpine complex 174 Figure 4.4.1 a: The effect of ferruginol on S Al 199B 176 Figure 4.4.2a: The effect of totarol on SAK3092 178 Figure 4.6: Isolated compounds 193 Figure 4.6.4a: Bail and stick model of formosanoxide 200 Figure 4.6.5a: Totarol analogue 416 202 Figure 4.6.5b: Totarol analogue 390 202 Figure 4.6.5c: Epicatechin gallate and epigallocatechin gallate 202 Figure 4.6.5d: Camosic acid 204 Figure 4.6.5e: Anacardic acid 205 Figure 4.6.5f: 5’-methoxyhydnocarpin 205 Figure 4.6.5g: Reserpine 206 Figure 5.1: Adult head louse 209 Figure 5.1.1a: Head louse egg 211 Figure 5.1.1b: Hatched egg (nit) 211 Figure 5.4.1 : Pisiferic acid and its congeners 227
- 1 0 - TABLE OF TABLES Page Table 2.1: MIC values (|ig/ml) for standard antibiotics tested against 78 S. aureus strains used in MIC and modulation assays Table 3.1: MICs (gg/ml) of cone extracts from Coniferae species 81 Table 3.2.1: MICs (pg/ml) of VLC extracts 82 Table 3.2.2: Yield of compounds isolated from C. lawsoniana 84 Table 3.3.1a: (500 Hz) and ^^C (125 Hz) NMR spectral data for CL-001 89 Table 3.3.1b: Long-range ^H—>^^C connectivities for CL-001 detected 90 in an HMBC experiment Table 3.3.2a: Long-range ^H^^^C connectivities for CL-003 detected 94 in an HMBC experiment Table 3.3.2b: ^^C (125 Hz) NMR spectral data for CL-002 and CL-003 95 Table 3.3.2c: *H (500 Hz) NMR spectral data (/in Hz) for CL-002 96 and CL-003 Table 3.3.3a: ^^C (125 Hz) NMR spectral data for CL-004 102 Table 3.3.3b: Long-range ^H—>^^C connectivities for CL-004 detected 103 in an HMBC experiment Table 3.3.4a: *H (500 Hz) and *^C (125 Hz) NMR spectral data for CL-005 108 Table 3.3.4b: Long-range ^H-^^^C connectivities for CL-005 detected 109 in an HMBC experiment Table 3.3.5a: Long-range ^H^^^C connectivities for CL-006 detected 112 in an HMBC experiment Table 3.3.5b ^H (500 Hz) and *^C (125 Hz) NMR spectral data for CL-006 114 Table 3.3.6a ^H (500 Hz) and *^C (125 Hz) NMR spectral data for CL-007 119 Table 3.3.6b Long-range ^H^^^C connectivities for CL-007 detected 121 in an HMBC experiment Table 3.3.7a: *H (500 Hz) and ^^C (125 Hz) NMR spectral data for CL-008 124 Table 3.3.7b: Long-range *H—>^^C connectivities for CL-008 detected 125 in an HMBC experiment Table 3.4.1: MICs (pg/ml) of C. nootkatensis VLC extracts 128 Table 3.4.2: Yield of compounds isolated from C. nootkatensis 129 Table 3.5.1a: *H (500 Hz) and ‘^C (125 Hz) NMR spectral data for CN-001 131 Table 3.5.1b: Long-range ^H^^^C connectivities for CN-001 detected 132 in an HMBC experiment Table 3.5.2a: ^H (500 Hz) and ^^C (125 Hz) NMR spectral data for CN-002 138 Table 3.5.3a: ^H (500 Hz) and ^^C (125 Hz) NMR spectral data for CN-003 142 Table 3.5.3b: Long-range ^H—>*^C connectivities for CN-003 detected 143 in an HMBC experiment Table 3.5.4a: ^H (500 Hz) and ^^C (125 Hz) NMR spectral data for CN-004 146 Table 3.5.4b: Long-range ^H^^^C connectivities for CN-004 detected 147 in an HMBC experiment Table 3.6.1: MICs (pg/ml) of cone extracts from Pinus species 151 Table 3.6.3a: ^H (500 Hz) and ^^C (125 Hz) NMR spectral data for PN-001 154 Table 3.6.3b: Long-range ^H-^^^C connectivities for PN-001 detected 155 in an HMBC experiment Table 4.1a: MICs (pg/ml) for standard antibiotics tested against 159 S. aureus strains used in MIC and modulation assays
-11 - Table 4. lb: MICs (p,g/ml) of isolated compounds against a standard ATCC 160 strain and effluxing strains of S. aureus Table 4.1c: MICs (pg/ml) of isolated compounds against EMRS A strains 161 Table 4.2a: MICs (pg/ml) of tetracycline, norfloxacin and erythromycin 165 in the presence and absence of isolated compounds Table 4.2b: MICs (pg/ml) of oxacillin in the presence of isolated 167 compounds Table 4.3.1a: MICs (pg/ml) for compounds in the presence and absence 171 of reserpine Table 4.3.2a: Reserpine and abietic acid resonances which showed a 172 change in chemical shift after mixing and overnight incubation Table 5.3.1: Results of initial screening of Chamaecyparis extracts 219 Table 5.3.2: Results of repeat pediculicide assay on selected VLC samples 220 Table 5.3.3: Repeat of pediculicide assay on C.L. VLC 5 and C.N. VLC 6 221 Table 5.3.4: Results for abietic acid and ferruginol in the pediculicide assay 222 Table 5.3.5: Activity of Chamaecyparis sesquiterpenes and diterpenes 222 against adult lice Table 5.3.6: Pediculicide assay on different concentrations of active 223 compounds
- 1 2 - ABBREVIATIONS
Calcium ions MHB Mueller-Hinton broth Cfu Colony forming units MIC Minimum inhibitory CHCI3 Chloroform concentration COSY Correlation spectroscopy MRSA Methicillin resistant Staphlyococcus aureus CPK Corey Pauling Kulton MTT 3-[4,5-dimethylthiazol-2- DEPT Distortionless yl]-2 ,5-diphenyl Enhancement by tétrazolium bromide Polarisation Transfer NaCl Sodium chloride DMAPP Dimethylallyl pyrophosphate NADH Nicotine amide adenine dinucleotide DMSG Dimethyl sulphoxide NCCLS National Committee for ECG Epicatechin gallate Clinical Laboratory EGCG Epigallocatechin gallate Standards EMRSA Epidemic methicillin NHS National Health Service resistant Staphylococcus NMR Nuclear magnetic aureus resonance EtBr Ethidium bromide NOESY Nuclear Overhauser effect Etc Ac Ethyl acetate spectroscopy FPP Famesyl pyrophosphate ppm parts per million GGPP Geranyl geranyl PTLC Preparative thin-layer pyrophosphate chromatography GPP Geranyl pyrophosphate QAC Quaternary ammonium compound HMBC Heteronuclear multibond coherence spectroscopy QSAR Quantitative structure activity relationship HPLC High performance liquid chromatography SAR Structure activity relationship HSQC Heteronuclear single quantum coherence see Staphylococcal cassette spectroscopy chromosome
H2SO4 Sulphuric acid TLC Thin-layer chromatography IPP Isopentenyl pyrophosphate VRSA Vancomycin-resistant MDR Multidrug-resistant Staphylococcus aureus MeOH Methanol VRE Vancomycin-resistant Mg^^ Magnesium ions enterococci
1 3 - ACKNOWLEDGEMENTS
My immense thanks go to my two supervisors Dr. Simon Gibbons and Professor Elizabeth Williamson, without whose support, encouragement and sharing of their expertise I would not have completed this PhD.
I am also very gratehil to Stiefel Research Laboratories for funding my studentship and particularly to Dr. David Small for all his interest and support.
My thanks also go to Dr. Caroline Priestley who started me off on my conifer studies and who organized the supply of cones for the initial screening; Chris Reynolds and Daniel Luscombe at Bedgebury Pinetum for further supply of conifer material; Neale Wareham at Stiefel for the GC-MS work; Professor Glenn Kaatz for the efflux inhibition assays and Ian Burgess and Maxine Harris at Insect R&D.
I would also like to thank the School of Pharmacy, Professor Michael Heinrich and all my colleagues in the Centre for Pharmacognosy and Phytotherapy, particularly Cory for keeping the lab running smoothly and my fellow PhD student Gemma for sharing her Liowledge of HPLC. I also thank Dr Mire Zloh for assistance with NMR and all the ‘behind the scenes’ people at SOP who have helped me during my time there.
A big thank you to my mother for all her support, although her contracting an MRSA infection was perhaps taking the concept of ‘support’ a little too far.
This thesis is dedicated to my father.
- 1 4 - 1.0 INTRODUCTION
1.1 THE PROBLEM OF RESISTANT BACTERIA
The emergence of bacterial pathogens with resistance to most antibiotics has become a health threat worldwide. The ability of many of these pathogens to evade antibiotic action and the rapid emergence of resistance to new antibiotics and the ability of resistance genes to pass from one bacterial species to another is of particular concern.
Nosocomial infection is a major problem in hospitals (Levy, 2005; Overbye and
Barrett, 2005) with the most common pathogens including methicillin-resistant
Staphylococcus aureus (MRSA), vancomycin resistant Enterococci (VRE), fluoroquinolone-resistant Pseudomonas aeruginosa and Streptococcus pneumoniae.
These virulent pathogens can cause serious conditions including bacteraemia, endocarditis, haemolytic pneumonia, infection of lungs, wounds and the urinary tract. Even bacteria which were once considered to be of no particular health risk have become multidrug-resistant (MDR). One example is the Gram-negative
Acinetobacter, species of which cause pneumonia and are now considered one of the most difficult nosocomial infections to treat (www.idsociety.org).
Bacterial infections result in a huge cost, not only in patient deaths, but they also impact considerably in the cost to health services in treating such infections. In the
United Kingdom, the National Audit Office reported that the annual cost to the
National Health Service (NHS) of nosocomial infection is approximately £1 billion and that such infections cause 5,000 deaths per year (www.nao.org.uk). The situation is as serious in the United States, and has been highlighted in a recent white paper, “Bad Bugs no Drugs”, by the Infectious Diseases Society of America
(www.idsociety.org). They estimate that in the U.S. two million people suffer from
15 - nosocomial infections each year, leading to 90,000 deaths and that the cost to society is nearly $5 billion per annum.
There has been a large decline in antibacterial research by major pharmaceutical companies (Overbye and Barrett, 2005; Projan, 2003). In a report by Spellberg et al.
(2004) it was stated that out of 506 drugs in late stage development by the world’s 15 largest pharmaceutical companies, only six are new antibacterial agents and they are all derivatives of existing antibiotics.
There is obviously an urgent need for new antibacterials with new modes of action.
High throughput screening, genomics and target-based drug design have not yet led to the hoped for success in developing new antibiotics. Natural product derived drugs are more likely to be able to cross the bacterial cell membrane, a stumbling block for target-based drug design. Despite numerous reports of good antibacterial activity for plant compounds, there are at present no plant-derived antibiotics in clinical use (Gibbons, 2004b). This is surprising, since plant compounds have been of considerable benefit in other areas of illness and disease, from one of the first commercial drugs, asprin, semisynthesised from salicylic acid found in Salix species, to more recent discoveries such as the anti-cancer drugs taxol from Taxus brevifolia and vincristine and vinblastine from Catharanthus roseus.
1.2 ANTIBACTERIALS
Antibacterials or biocides tend to have broad-spectrum activity, mainly targeting cell surface components, whereas antibiotics are more specific in their mode of action, often with their target site within the cell and they usually have a narrower spectrum of activity. (Gilbert and McBain, 2004; Cole et al., 2003).
- 16 - Antibacterials have been used as disinfectants and antiseptics and include quaternary ammonium compounds (QACs), chlorhexidine and products containing oleoresin, pine oil and coal tar. Polymyxins have also been used as antibacterials. They are basic cyclic polypeptides and act as detergents, disrupting the bacterial membrane.
1.2.1 Triclosan
One of the most widely used biocides is
triclosan. It has been incorporated into a
myriad of products including cleaning
materials, washing-up liquids, plastics for food containers and chopping boards, toys, cosmetics, deodorants and shampoos.
Concern was raised over the use of triclosan on the discovery that it targets the enoyl-acyl carrier protein reductase enzyme of Escherichia coli which is involved in fatty acid synthesis and is encoded by the fa b i gene. (Levy et al., 1999). This enzyme is also the target of isonaizid, a drug used against tuberculosis and there was concern that the use of triclosan might select for mutants of this gene. There are no reports of in vitro resistance to triclosan in any other organism other than E. coli, and
Mycobacterium tuberculosis species which are resistant to isonaizid remain sensitive to triclosan (Gilbert and McBain, 2004). It has been suggested that most bacteria have targets other than enoyl reductase which are much more sensitive to triclosan and therefore the resistance seen in E. coli has not developed in other species of bacteria. However reduced susceptibility to triclosan has been detected in EMRSA-
16 clinical isolates (Brenwald and Fraise, 2003).
- 1 7 - A study by Cole et al. (2003) on clinical and environmental samples collected from
the homes of users and nonusers of antibacterial products found more potential
pathogens in the homes of nonusers. They did not detect any cross-resistance
between antibacterials and antibiotics. None of the isolates of S. aureus showed
resistance to oxacillin or vancomycin and no resistance by E. coli or Klebsiella pneumoniae to the cephalosporins ceftazidime or cefotaxime was detected. These
authors concluded that their results refute fears that overuse of antibacterial products
in the home promotes the development of resistant bacteria and state that targeted use
of such products will reduce the prevalence of potential pathogens in the home. This
conclusion is disputed by others who believe that it may be only a matter of time
before widespread resistance to triclosan arises (Levy, 2005). The Food and Drug
Administration in the United States is sufficiently concerned to have ordered an
investigation into the use of biocides in antibacterial consumer products (Hileman,
2005).
The results reported by Cole et al. (2003) should be treated with some caution, as
although samples were taken from the hands and mouths of individuals from the
homes used in the study, the results do not reveal what effect, if any, the use of
biocides such as triclosan have on human gut flora. When included in products such
as soaps, toothpastes, handcreams and washing-up liquids, some residue of triclosan
is likely to reach the gut, where there is the possibility that it might select for
resistant bacteria, or if nothing else, potentially upset the balance of gut flora.
- 1 8 - 1.2.2 Silver Ions
Silver ions (Ag ) are known to have antibacterial activity and silver impregnated dressings such as Acticoat’’^’'^ and Contreet'^'^ are in current clinical use to treat wound infections. It is also possible to buy silver impregnated plasters in chemists, although there is concern that overuse of such plasters and dressings will lead to resistance (Percival et a l, 2005). Such topical application of an antibacterial is also used in combinatory therapy, whereby the patient is given antibiotics systemically and antibiotics or antibacterials are applied topically. In the picture shown below, the elderly patient’s leg had an area of eczema which became ulcerated and infected with MRSA. Treatment was with a combination of two antibiotics (rifampicin and fucidin) taken orally, and application of a silver impregnated dressing which was changed every three days. The treatment was successful and the leg healed.
However, it was suggested that the silver ions were solely responsible for the patient’s recovery and that the antibiotics administered were unlikely to have been effective against MRSA (Dr. David Newman, National Cancer Institute, USA, personal communication).
Figure 1.2.2:Treatment of an MRSA infected leg ulcer with silver dressing
- 19 - 1.3 ANTIBIOTICS
The heyday of antibiotics was from the 1950’s to the mid 1980’s (Leeb, 2004;
Amyes, 2001), when it appeared that antibacterial infections had been brought under control by antibiotics and once life-threatening conditions such as tuberculosis and bacteraemia could be cured. Unfortunately, bacterial resistance has caught up with and even overtaken antibiotics in current clinical use and the lack of new antibiotics with new structures and modes of action is of considerable concern (Levy, 2005;
Walsh and Wright, 2005).
1.3.1 p-lactams p-Lactam antibiotics are based on the natural product penicillin and include the penicillins, cephalosporins and carbapenems. They all have the characteristic P- lactam ring, but vary in the type of ring fused to the P-lactam and the composition of the side chain. For example, the cephalosporins have a six membered ring fused to the p-lactam ring, which provides greater scope for modification of the molecule.
Third generation cephalosporins - cefotaxime and ceftazidime have been developed to improve activity against Gram-negative bacteria. Fourth generation cephalosporins, including the zwitterionic cefepime and ceQ)irome are able to penetrate the cell wall much faster than other cephalosporins.
Carbapenems have a five-membered ring attached to the lactam ring but, unlike the penicillins, the sulphur atom is replaced by a carbon. They are broad-spectrum against Gram-positive and Gram-negative bacteria and, being zwitterionic, are able to penetrate the cell wall quickly. The drawback with the carbapenems is that they are expensive; imipenem has to be injected and there are some problems with
2 0 - toxicity. Meropenem was developed more recently and has fewer problems with toxicity and is in clinical use in hospitals (Aymes, 2001). o II u H Figure 1.3.1: p-Lactam ring with a fused thiazolidine ring as seen in the penicillins
CO2H
The P-lactams are broad-spectrum antibiotics and have been successfully used against both Gram-positive and Gram-negative bacteria. They act as inhibitors of the final step of bacterial cell wall synthesis in which the peptidoglycan strands are cross-linked. This cross-linking is via a pentapeptide side chain with two terminal
D-alanine residues and is carried out by a D-alanyl-D-alanine transpeptidase enzyme.
The p-lactams inhibit the transpeptidase enzyme, forming a stable ester between the lactam and the serine residue at the enzyme’s active site (Livermore and Williams,
1996). These transpeptidase enzymes are also referred to as penicillin binding proteins (PBPs). One mode of resistance to P-lactams is by modification of the
PBPs, as seen in some Gram-negative species. Another form of modification is by acquisition of a gene encoding a modified PBP, as seen in MRSA srains of S. aureus whereby acquisition of the mecA gene which encodes a low-affinity PBP, confers resistance to methicillin and to all p-lactams. Another mode of resistance is by inactivation of the P-lactam antibiotic by P-lactamase enzymes which hydrolyse the p-lactam ring. Of some cause for concern are the metallo-p-lactamases which have a different mechanism of hydrolysis from the serine P-lactamases and are active against the new generation cephalosporins and carbapenems. At present, these
-21 metallo-P-lactamases are confined to opportunistic, clinically rare Gram-negative species, but the concern is that they may be acquired by more widespread virulent species (Fisher et al., 2005).
1.3.2 Glycopeptides
The glycopeptides vancomycin and teicoplanin also act to inhibit cell wall synthesis, but by a different mode of action from the P-lactams. Vancomycin binds to peptidoglycan cell wall precursors containing a terminal D-alanine-D-alanine moiety.
Binding inhibits the transglycosylation and/or transpeptidation steps thereby preventing the polymerisation of peptidoglycan. Vancomycin and teicoplanin are used clinically to treat Gram-positive bacterial infections that have proved resistant to antibiotic therapy (Kahne et al., 2005). They have proved particularly effective against Enterococcus species which cause endocarditis and in treating MRSA infections. The use of vancomycin is restricted due to its toxicity and because it is often a ‘last resort’ against MRSA when all other treatment has failed. Teicoplanin is a lipoglycopeptide with a fatty acyl chain which makes it more lipophilic than vancomycin.
Vancomycin resistance first appeared in the enterococci by acquisition of a five gene cassette VanA which confers resistance to both vancomycin and teicoplanin or VanB which only results in vancomycin resistance. Two of these genes (yanR and vanS) encode proteins that form a regulatory system, leading to transcription of the
VanHAX genes. The VanHAX proteins are enzymes which lead to modification of the glycopeptide target. This results in cell wall peptidoglycan precursors with a D- alanine-D-lactate moiety at the end of the peptide stem, but which can still function as substrates for the transpeptidase PBPs (Kahne et al., 2005). Although vancomycin
- 2 2 - intermediate resistant S. aureus had previously been reported, the first instance of vancomycin-resistant S. aureus (VRSA), where the clinical isolate was found to contain the VanHAX genes, was described by Chang et al. in 2003.
1.3.3 Macrolides
The macrolide antibiotics act by inhibiting protein synthesis. They bind to the 50S subunit of the bacterial ribosome, but not to the mammalian ribosome, which is of considerable value for clinical use. Erythromycin is a polyketide which contains a
14 atom macrocyclic ring. It was the first macrolide to be used clinically in 1952, against S. aureus infections, and erythromycin resistance was first reported in the same year (Katz and Ashley, 2005).
Resistance to erythromycin may be via several routes, one method being the acquisition of one of the erm genes encoding an erythromycin resistance methylase.
Méthylation of the 23 S rRNA is thought to prevent erythromycin binding to the ribosome (Poole, 2001). Another mode of resistance is via an efflux pump, an example being the macrolide specific MsrA pump of S. aureus.
HO
OH
Figure 1.3.3: Erythromycin
2 3 - Erythromycin has poor bioavailability and has unpleasant side-effects causing gastric upsets. Second generation macrolides including clarithromycin and azithromycin
were developed from erythromycin. They are more lipophilic and demonstrate
improved bioavailability and increased half-life. Both compounds are in clinical use
and clarithromycin is used to treat gastric ulcers cased by the bacterium Heliobacter pylori and to treat AIDS-related respiratory infections (Katz and Ashley, 2005).
Third generation macrolides are in development due to the pressure caused by macrolide resistant S. pneumoniae. Telithromycin is prepared from clarithromycin
and is in clinical use in Europe (Katz and Ashley, 2005).
1.3.4 Tetracyclines
Tetracycline antibiotics target bacterial protein synthesis. They bind to the bacterial
ribosome inhibiting aminoacyl transfer RNA from binding at the A site on the
ribosome. The genes for resistance are encoded by tet determinants which are
usually carried on a plasmid or transposon (Poole, 2001). There are several ways in
which bacteria can show resistance to tetracyclines: efflux (TetA, TetK, TetL),
ribosomal protection (TetM) and antibiotic inactivation (TetX).
\ / HQ .OH
NH
OH OHOH
Figure 1.3.4: Tetracycline
- 2 4 - 1.3.5 Fluoroquinolones
Fluoroquinolones target the enzymes DNA gyrase and topoisomerase IV involved in unwinding the DNA helix at replication. Resistance is often in the form of target site mutations in the DNA. Newer fluoroquinolones including sitafloxacin and
clinafloxacin have been developed which target both the gyrase and topoisomerase
IV enzymes and therefore mutations have to arise in both genes to confer resistance
on the bacterium (Poole, 2001). Another mode of resistance to fluoroquinolones is
by efflux. In S. aureus, fluoroquinolones, including norfloxacin and ciprofloxacin
are substrates for the multidrug-resistant pump NorA. However, newer
fluoroquinolones have been developed (sparfloxacin and trovafloxacin) which are poor substrates for the NorA pump (Markham et al., 1999).
O
HN
Figure 1.3.5: Norfloxacin
- 2 5 - 1.4 METHODS OF BACTERIAL RESISTANCE
There are many methods by which bacteria are known to exert their resistance to antibiotics:
• Increased impermeability of cell wall or membrane
• Efflux of antibiotic
• Modification of target site
• Inactivation of antibiotic
• Ribosomal protection
• Overexpression of target protein
Resistance may be intrinsic, or acquired. For example, possession of an efflux pump is intrinsic resistance if it forms part of the normal complement of genes for a particular bacterium, but if the gene(s) encoding the pump are on a plasmid then it is likely that the pump was acquired from another bacterium or from the environment.
1.4.1 Intrinsic resistance
Intrinsic resistance includes structures and processes which are part of the normal make-up of a bacterium. Over-expression of the target is seen in bacteria which over-express PBPs or P-lactamases. Reduced susceptibility to methicillin has been reported for some S. aureus strains which do not possess the SCCwec element and this has been attributed to over-expression of penicillinase, or to mutations in the
PBPs leading to reduced affinity for P-lactam antibiotics. Chromosomally encoded efflux pumps, such as NorA are part of the normal complement of genomic DNA, however, increased resistance to an antibiotic may occur due to over-expression of the pump as seen with strain SAl 199B (Kaatz et al., 1993) or if the genome contains
- 26 - multiple copies of the gene(s) encoding the pump. Thickening of the cell membrane
or the reduction in or modification of cell wall porins in response to antibiotic attack
is also another form of intrinsic resistance.
1.4.2 Acquired Resistance
Acquired resistance can be detected whereby resistant strains possess proteins which
are not present in the wildtype susceptible strain. There are several means by which
‘foreign’ DNA can be acquired by bacteria. Bacterial conjugation is one method by
which new genetic material can be passed from one bacterium to another. Resistance
genes may be carried on a plasmid which can be passed between bacteria of the same
species or between bacterial species. Plasmid-encoded class A p-lactamases have been transmitted from Gram-positive to Gram-negative bacteria. The most common plasmid-encoded resistance gene found in clinical Gram-negative bacteria is TEM-1
which encodes a class A P-lactamase (Amyes, 2001).
Transposons, also called ‘jumping genes’ are mobile genetic elements. Insertion
sequences can move plasmid or chromosomal DNA from one bacterium to another.
A piece of the DNA strand between two insertion sequences is mobilised, if this is a
resistance gene, it will be transposed to another bacterium. This event occurs
directly after DNA replication, so that the donor bacterium does not lose the genetic material.
The tet determinants encoding resistance to tetracycline are usually plasmid or
transposon encoded, and therefore tetracycline resistance can spread to a range of bacterial species (Marshall et al., 1997). One of the most well reported examples of
acquired resistance is the acquisition of the mecA gene which leads to methicillin- resistance in strains of S. aureus (Hiramatsu et al., 2001).
- 2 7 - 1.4.3 Efflux Pumps
Removal of antibiotics and other toxic compounds from the cell is a major mode of resistance in both Gram-positive and Gram-negative bacteria. Efflux pumps may be specific for one type of compound, such as the MsrA pump of S.aureus which effluxes macrolides, or they may be multidrug-resistant with many substrates, for example, the NorA pump, also expressed by S. aureus which effluxes fluoroquinolones, quartemary ammonium compounds and many structurally unrelated compounds (Markam et al., 1999) and the multidrug transporter AcrB from
E. coli. which effluxes P-lactams, erythromycin, tetracycline, detergents, ethidium bromide, dyes and many other compounds (McKeegan et al., 2004). Efflux pumps also function in the uptake of nutrients and ions (Li and Nikaido, 2004).
Multidrug-resistance by means of an efflux pump is utilised by organisms other than bacteria; examples include the CaMDRl pump of Candida albicans or the PfMDR of Plasmodium falciparum (Li and Nikaido, 2004). The mammalian MDR p- glycoprotein is found in tumours and effluxes chemotherapy agents (Qian et al.,
2005).
In bacteria there are five families of efflux transporters. Two large superfamilies comprise the ATP-binding cassette (ABC) and the major facilitator (MFS) families, and three small families: the multidrug and toxic compound extrusion (MATE) family; the small multidrug-resistance (SMR) family and the resistance-nodulation- cell division (RND) family (Li and Nikaido, 2004). The ABC transporters use the energy from ATP hydrolysis to drive efflux, whereas the other pump families utilise
- 2 8 - a proton antiporter mechanism in which influx of protons is coupled to efflux of compounds from the cell (Marshall and Piddock, 1997).
Efflux pumps from the major facilitator superfamily have 12 membrane-spanning helices. From crystallography and other structural studies, it has been suggested that each transmembrane domain of an efflux pump is capable of drug translocation
(McKeegan et al., 2004). This could explain why some pumps have multiple structurally dissimilar substrates as different transmembrane domains could have different active sites, in terms of size, conformation and presence of charged or neutral amino acid residues.
The efflux pumps of Gram-positive bacteria are simpler in structure than those of
Gram-negatives. In Gram-positive bacteria compounds can be effluxed across the bacterial cytoplasmic membrane from where they will diffuse across the cell wall into the medium. For Gram-negative bacteria however, compounds are effluxed across the cell membrane into the periplasm and they are then transported across the outer membrane into the environment by a tripartite transport complex; this is seen with the AcrB MDR pump of E. coli. AcrB belongs to the RND family and is an inner membrane protein, TolC is an outer membrane protein and the two proteins interact by a membrane fusion protein AcrA forming a tripartite complex (McKeegan et al., 2004).
- 2 9 - ENVIRONMENT
Outer membrane
TolC AcrA
NorA Cell membrane AcrB
Drugs Drugs
CYTOPLASM
Figure 1.4.3: Drug efflux in Gram-positive and Gram-negative bacteria
NorA {S. aureus) MF - major facilitator superfamily AcrB {E. coli) RND - resistance-nodulation-cell-division AcrA - membrane fusion protein TolC - outer membrane protein (after Paulsen, 2003)
1.4.4 S. aureus Efflux Pumps
The S. aureus NorA MDR pump belongs to the MF superfamily. The pump effluxes a range of structurally unrelated compounds, including norfloxacin, ciprofloxacin,
QACs, dyes and ethidium bromide (Markham et al., 1999; Marshall and Piddock,
1997). NorA is a close homologue of the Bmr MDR pump of Bacillus subtilis. A new S. aureus efflux pump, NorB, has recently been reported, which also effluxes fluoroquinolones and EtBr (Truong-Bolduc et al., 2005).
- 3 0 - The QacA and QacB pumps are plasmid encoded. They also belong to the MF superfamily but have 14 rather than 12 transmembrane helices. QacA and QacB, both efflux the same substrates: EtBr, chlorhexidine, rhodamine and other dyes and
QACs, but QacB has reduced affinity for chlorhexidine (Marshall and Piddock,
1997).
The MsrA pump is an ABC transporter found in Staphylococci species. MsrA confers resistance to macrolides including the 14-membered erythromycin and clarithromycin and 15-membered azithromycin and streptogramin B (Li and Nikaido,
2004).
TetK is a member of the MF superfamily of transporters. Tetracycline is complexed with divalent cations such as Mg^^ and effluxed (Marshall and Piddock, 1997).
31 - 1.5 STAPHYLOCOCCI
Staphylococci are Gram-positive aerobic bacteria which are spherical in shape.
Staphylococcus aureus and the coagulase-negative Staphylococcus epidermidis are common components of human skin flora, although S. aureus is often found in the nasal passages. Staphylococci do not pose a problem to a healthy host, but serious problems can arise if they gain entry to the body. This may be via a simple cut or skin lesion or a patch of eczema, or infection may occur in a wound or bum or at the site of an operation (Amyes, 2001). S. aureus is considered the more serious pathogen, although S. epidermidis can be a particular problem, forming biofilms infecting surgical implants. Staphylococci express the cell surface proteins laminin and fibronectin which facilitate adhesion to host tissue.
Figure 1.5a: S. aureus
Figure 1.5b: S. aureus biofilm
(Photographs from www.lib.uiowa.edu)
- 3 2 - s. aureus infections can cause serious conditions such as bacteraemia, endocarditis, haemolytic pneumonia, and infection of the lungs, wounds and urinary tract. Many
strains including MRSA strains also secrete exotoxins such as a-toxin which causes leakage of cell contents and septic shock. Staphylococcal enterotoxins cause food poisoning, and toxic shock syndrome toxin (TSST-1) the causative agent of toxic
shock syndrome can be fatal. Panton-Valentin leukocidin (PVL) is a toxin secreted by many community-acquired strains of MRSA. It causes lysis of white blood cells leading to necrotising pneumonia and children and young people are particularly
susceptible (Gillet et al., 2002). The PV~luk gene is carried on a bacteriophage which can inject the DNA into a host cell. The genes for other toxins, including toxic shock syndrome and enterotoxins B and C, can be carried on S. aureus pathogenicity islands which are transported by helper phages and move at high frequency (Lindsey and Holden, 2004).
1.5.1 MethiciUin-resistant S. aureus (MRSA)
Methicillin is a semi-synthetic penicillin, resistant to staphyloccoal penicillinase enzymes. It was first used clinically in 1960 and the first report of a methicillin- resistant S. aureus isolate was just one year later, in 1961 (Jevons, 1961).
OMe Figure 1.5.1a: Methicillin
OMe
- 3 3 - Methicillin-resistance is conferred by acquisition of the mecA gene which is carried on a mobile genetic element called the staphylococcal cassette chromosome (SCC).
The SCC is a large DNA element of 21 - 67 kb, along with mecA it carries recombinase genes {ccrA and ccrE) which are involved in the integration of mecA into the bacterial chromosome. Some SCCs also carry the regulatory genes m eci and mecRl which encode a repressor protein and a signal transduction protein respectively. SCCs may possess genes which confer resistance to non-P-lactam antibiotics, which is one of the factors that make MRSA such a virulent pathogen.
There are four types of SCC; type I to III are associated with clinical isolates and type IV is associated with community-acquired MRSA (Lindsay and Holden, 2004;
Hiramatsu et a/., 2001). Type I SCCmec does not possess any antibiotic resistance genes other than mecA and was found in MRSA isolates in the I960’s. Types II and
III which appeared in the 1980’s have multiple antibiotic resistance genes (Hiramatsu et a/., 2 0 0 1 ).
It has been suggested S. aureus acquired the mecA gene from other staphylococcal
species. MecA has been detected in nine coagulase-negative staphylococcal species, including S. epidermidis and S. haemolyticus (Suzukiet al., 1992) and Hiramatsu et al. (2001) reported that Class IV SCCmec is often seen in community acquired strains of S. epidermidis.
Epidemic strains of MRSA (EMRSA) have become established in hospitals across the world. In the UK, EMRSA-15 and EMRSA-16 were the major strains found in
MRSA bacteraemia isolates in a study from 26 hospitals (Johnson et al., 2001). In
2002, a new epidemic strain, EMRSA-17, was reported (Aucken et al., 2002). As
-34 well as being resistant to multiple antibiotics, this strain also demonstrated borderline resistance to the glycopeptide teicoplanin.
Community-acquired MRSA is now prevalent in many countries (Zetola et al.,
2005). It can be carried by healthy individuals and therefore may go undetected for long periods of time. In the United States, there has been a large increase in the number of outbreaks of community-acquired MRSA among sports players in colleges and also amongst professional players (www.idsociety.org).
Since 2001, all acute NHS Trusts in England have been required to report instances of MRSA bacteraemia to the Department of Health’s MRSA surveillance scheme
(www.dh.gov.uk). The annual figures for reported MRSA bacteraemias had shown a
consistent increase for the first three years, but the figures for 2004/05 indicated a
6.1% reduction in cases at 7,212 compared with 7,684 in 2003/04. The Government has claimed success for measures it introduced to combat MRSA in hospitals.
However, the figures are disputed by the MRSA support group who suggest that
instances of MRSA bacteraemia are much higher but many cases do not get reported
or included in the figures (www.mrsasupport.co.uk). MRSA has become a highly
emotive issue in the UK, with prominent reporting of the latest bacteraemia figures,
in all areas of the media. It became a political issue at the 2005 UK General Election
with the major political parties claiming they would solve the problem of hospital
acquired infection. The British public is very aware of the problem of MRSA probably due to some sensational reporting in the press “MRSA turns into flesh bug” being a prime example. However, there was also the tragic case in 2005 whereby a
- 3 5 - one-day-old healthy baby contracted MRSA in a maternity unit in Ipswich and died from the infection.
\v» Hospital chiefs admit they cannot stop killer superbug
Mm m of hôpital tfusl MRSA kills one-day-old baby |
Figure 1.51.b: Headlines from selected articles on MRSA (2004/05) from the London Evening Standard
There are disputes about the best way to try to prevent the spread of MRSA in hospitals, however the general consensus of opinion seems to favour regular Slop! handwashing above all else (Edwards, 2004). The NHS
‘cleanyourhands’ campaign is aimed at hospital staff,
V ® / honds patients and visitors. Patients are now encouraged to insist that healthcare workers clean their hands between treating one patient and the next. There are also disagreements about how MRSA is contracted; it is usually considered that MRSA is a mainly nosocomial infection that patients contract whilst in hospital being treated for a different condition. However, some hospitals have blamed patients and even visitors for bringing the infection into
- 3 6 - the hospital in the first place and then passing it on to others (Edwards, 2004). The origin of the infection could prove very difficult to trace, since an infected patient may bring MRSA into the hospital, but unless it is a community acquired strain, they could have acquired the infection from a previous hospital stay at the same hospital or at a different one. One possibility is to screen patients for MRSA and, if infected, treat them before admission. There are also discrepancies with this method, as some hospitals screen all patients whilst others only screen high risk patients or those known to have been previously infected (Marshall et al., 2004). A common site of colonisation is the nasal passages and the usual treatment is with a mupirocin nasal
spray and bathing with a soap solution containing chlorhexidine or triclosan, however this is not always successful (Marshall et al., 2004).
1.5.2. Penicillin Binding Protein (PBP2')
The mecA gene encodes a lower affinity penicillin binding protein (PEP) PBP2’, also denoted as PBP2a. This protein was identified by Hartman and Tomasz (1984) as being present only in MRSA and not in methicillin-sensitive strains. The S. aureus
PBPs 1-4 are inactivated by P-lactam antibiotics. In MRSA PBP2' performs the transpeptidase function of the other PBPs. However, unlike PBP2, it does not have transglycosidase activity. PBP2 performs a transglycosidase step whereby the cell wall precursors, A-acetylglucosamine and A-acetylmuramic acid are linked, and this is followed by the transpeptidase step in which the penultimate D-alanine on the peptide is linked to a glycine on the newly formed cell wall. PBP2' therefore has to use the transglycosidase domain of PBP2 to build the wall. PBP2’ also requires
specific cell wall precursors which contain a pentaglycine side chain attached to the
lysine residue at position 3 of the peptide stem (Guignard et al., 2005). This requires
- 3 7 - the participation of accessory proteins including the chromosomally encoded FemA,
FemB and FemX (also known as FmhB) which are responsible for the attachment of the five glycine residues to the position 3 lysine. FemX catalyses the transfer of the first glycine amino acid and FemA and B are involved in the attachment of the remaining four glycines, as two separate glycylglycine dipeptides (Mallorqui-
Femandez et al., 2004).
A-acetyl- A-acetyl- E glucosamine muramic acid
femC /.-A lan in e
/)-Glutamine-NH? femX
L-Lysine-Œy-Gly-Gly-Gly-Gly A A
/)-Alanine
femA femB /)-Alanine
Figure 1.5.2: The precursors required for PBP2' activity
PBP2 performs the transglycosidase step, linking N- acetylglucosamine to an A-acetylmuramic acid molecule in the growing cell wall. Proteins encoded by femA, femB, femC and femX are required for the full expression of methicillin resistance. Inactivation of any of the genes results in decreased resistance to methicillin even though PBP2' is present. (From: Guignard et al., 2005)
- 38 - It is interesting that the older penicillins such as penicillin G, amoxicillin and ampicillin have more than a 10-fold greater affinity for PBP2' than methicillin exhibits (Guignard et al., 2005). This led to the development of p-lactams with an affinity for PBP2' that is 100-fold greater than that of methicillin. These molecules are also resistant to hydrolysis by penicillinase enzymes. Two of these compounds, the carbapenem CS-023 and the cephalosporin ceftobiprole and its pro-drug ceftobiprole medocaril are now in Phase 11 and Phase 111 clinical trials (Guignard et a l, 2005).
1.6. NEW ANTIBIOTICS
In June 2005, the FDA granted approval to a new antibiotic tigecycline, marketed as
Tygacil by Wyeth. Tigecycline is a glycylcycline which binds to the bacterial ribosome, interfering with protein synthesis. It will be prescribed for treatment of
MRSA and VRE infections (Senior, 2005).
Synercid approved in 1998 is a combination of the natural products quinupristin and dalfpristin (pristinamycins 1 and 11). Synercid acts by blocking protein synthesis in the elongation phase (Kahne et a l, 2005).
Linezolid, approved in 2000, is a synthetic oxazolidinone which acts on protein
synthesis by binding to the P site at the 50S subunit of the bacterial ribosome (Meka and Gold, 2004).
O
Figure 1.6 N Linezolid
- 3 9 - Resistance to Linezolid has already been seen in clinical isolates, in the form of single nucleotide mutations in genes encoding the 23 S ribosomal RNA (Meka and
Gold, 2004).
Daptomycin is a lipopeptide and was approved by the FDA in 2003 as Cubicin. It disrupts the bacterial membrane, resulting in ion leakage leading to death of the bacterial cell (Kahne et al., 2005). Second generation glycopeptides, oritavancin and dalbavancin are in development for activity against VRE, MRSA and VRSA (Kahne et al., 2005).
40 1.7 MODULATION
A modulator is a compound which potentiates the antibacterial activity of another compound, usually an antibiotic. There are many potential benefits of using a modulator together with an antibiotic:
• It may restore antibiotic sensitivity in a resistant strain.
• Antibiotics which are no longer effective may be re-introduced.
• Less antibiotic may be used, reducing side effects and toxicity.
• Delay in emergence of antibiotic resistance.
It has been shown that that this type of combinatory therapy delays the development of antibiotic resistance (Markham et al., 1999). A modulator of p-lactam resistance has been used clinically for over 20 years. Augmentin'^^ is a combination of the antibiotic amoxicillin and the P-lactamase inhibitor clavulanic acid which binds the p-lactamase more quickly and more tightly than the penicillins, and is mainly used to treat respiratory tract infections. Timentin^'^ is a combination of clavulanic acid and ticarcillin and is used to treat severe hospital infections caused by P-lactam-resistant bacteria (Amyes, 2001). Other p-lactamase inhibitors in clinical use are sulbactam and tazobactam. Sulbactam is combined with ampicillin, produced as Unasyn"^^ and the product Tazocin^^ is a combination of tazobactam and piperacillin
Of considerable interest are the polyphenols isolated from green tea, particularly epicatechin gallate. This compound has been shown to potentiate oxacillin activity against MRSA by up to 500-fold (Stapleton et al., 2004; Shiota et al., 1999).
Electron microscopy studies by Hamilton-Miller and Shah (1999) showed that a compound from tea (denoted compound P) disrupted cell division and separation of daughter cells and caused thickening of the cell wall in MRSA, but methicillin-
-41 sensitive cells were unaffected. The related compound epigallocatechin gallate was
shown to potentiate the activity of the anti-cancer drug doxorubicin in cells
expressing P-glycoprotein mediated multidrug-resistance (Qian et al., 2005).
Epicatechin gallate and epigallocatechin gallate are discussed further in Section
4.6.5).
Modulation activity also includes efflux pump inhibition and this has been demonstrated by the plant alkaloid reserpine which inhibits the S. aureus pumps
NorA and TetK (Markham et al., 1999). Unfortunately, reserpine is neurotoxic at the
concentrations required for efflux pump inhibition and cannot be used clinically.
Stermitz and co-workers (2000) isolated the flavonolignan 5’-methoxyhydnocarpin
from Berberis fremontii which inhibited efflux of berberine. The same group has
shown that many plant metabolites had significantly enhanced antibiotic activity in the presence of an MDR inhibitor (Tegos et al., 2002). They also suggested that plant metabolites should not be considered inactive until they have been assayed in
the presence of an efflux pump inhibitor.
Other resistance modifying plant compounds include anacardic acid (Kubo et al.,
1992) and camosic acid (Oluwatuyi et al., 2004).
- 4 2 - 1.8 CONIFERS
Gymnosperm means ‘naked-seed’ and the seeds of coniferous gymnosperms are usually exposed on scales on the surface of the cone, not contained within a protective fruit as seen in the angiosperms. The Coniferae is the largest division of
gymnosperms and this group is referred to as conifers, meaning cone producing.
However, not all conifers bear true cones. The typical cone is woody and resinous, produced by conifers of the Pinus genus, but some species of other genera produce modified cones which appear more like fruits, examples include the fleshy aril which
surrounds the seed of Taxus species, or the berry-like cones home by species of
Juniperus, or the fruits of Cephalotaxus (Vidakovic, 1991). There are seven families of conifers, some with only very few genera, such as the Cephalotaxaceae or the
Araucariaceae, which has three genera and includes the Monkey Puzzle tree. The largest conifer families are the Pinaceae which includes the firs, larches and pines, the Cupressaceae, containing the cypresses and junipers and the Podocarpaceae.
Other smaller families are the Taxaceae (yews) and Taxodiaceae which includes the
Swamp Cypress and Redwoods.
1.8.1 Chamaecyparis
The genus Chamaecyparis is in the Cupressaceae or Cypress family and the trees are
often referred to as ‘false cypresses’. The trees are usually conical in form with
scale-like needles which give the shoots a fem-like appearance. The small globose
cones (8-10 mm) become woody, ripening in the first year (Vidakovic, 1991). The
genus contains six species and numerous cultivars. The species are native to North
America or East Asia, but are grown in many other countries as garden and
4 3 - ornamental trees. Some species of Chamaecyparis may live for hundreds, or even thousands of years (Kuo and Chan, 2005; Vidakovic, 1991).
1.8.2 Chamaecyparis lawsoniana
Chamaecyparis lawsoniana (Murray) Parlatore, also known as Lawson’s Cypress or
Port-Orford cedar, is a native tree of North America. It is found in coastal and mountainous regions of southwest Oregon and northern California. It exhibits ffeezing-toleranee of temperatures as low as -25°C and will grow in many soil types, but does require a moist atmosphere and its inland range is limited to 70 km
(Vidakovic, 1991).
Figure 1.8.2b: C. lawsoniana shoot Figure 1.8.2a: C. lawsoniana http://web.reed.edu/trees
1.8.3 Chamaecyparis nootkatensis
Chamaecyparis nootkatensis (D. Don) Spach., also known as Nootka Cypress or
Alaska Cedar is also native to the Pacific Coast region of North America, stretching from Alaska to Oregon. It is found at sea level and at altitudes up to 2,100 m. Like
- 4 4 - c. lawsoniana, it is freezing-tolerant and requires a moist atmosphere. Its shoots tend to droop at the ends giving the tree a pendulous appearance. The cones which have four to six hooked scales, may ripen in September of the first year, but in cold regions they will not mature until April or May of the following year (Vidakovic,
1991).
Immature cones
Figure 1.8.3: C. nootkatensis shoot
C. lawsoniana and C. nootkatensis are very similar in appearance, but there is speculation over whether C. nootkatensis belongs in the Cupressaceae at all (Little et al., 2004). The ancient sciences of botany and plant taxonomy have, until recently, been based on observation and determined by what the eye could see, sometimes with the aid of a microscope. Present day taxonomy is determined by what the molecular biologist detects in terms of DNA sequences of living and fossil specimens.
- 4 5 - Other species of Chamaecyparis comprise C. formosensis Matsumura, called Taiwan red cypress due to the reddish colour of its bark (Hsu et a l, 1995). It is found in high mountainous areas of Taiwan. The wood is used in buildings and for expensive furniture (Linet a l, 1999). C. obtusa (Sieb. et Zucc.) Endl. is native to Taiwan and
Japan and the wood is also used in building material (Kuo and Chan, 2005). C. pisifera is indigenous to Japan and C. thyoides is found in the United States along the
Atlantic Coast, stretching from Maine to Florida and Mississippi (Vidakovic, 1991).
1.8.4 Ethnobotany ofChamaecyparis
There is very little reported for the use of Chamaecyparis species in traditional medicine. The Salish people of British Columbia consider that illness can result from inhaling the strong odour of Chamaecyparis (Turner, 1988). However, all
Chamaecyparis species have hard, aromatic wood which is highly prized. C nootkatensis was used by Native American peoples to make canoe paddles and it is still used to construct poles, paddles and boats (http://web.reed.edu/trees). The essential oil from the heartwood of C. nootkatensis was shown to have activity against anaerobic bacteria and yeast, including Fusobacterium necrophorum and
Clostridium perfringens which cause foot diseases and other infections in domestic animals (Johnston et al., 2001). It was suggested that wood shavings of C. nootkatensis, used as bedding might help reduce these infections.
In Japan, the wood of C. obtusa is valued for use in construction of important buildings such as temples and shrines and is also considered to have hygienic properties for use as counter tops in sushi bars (Koyama et al., 1997).
There are also many reports in the literature on the activity of Chamaecyparis species against termites, ticks and other insects (see Section 5.1).
- 46 - 1.8.5 Chemistry ofChamaecyparis
The major components which have been extracted from species of Chamaecyparis
are monoterpenes in essential oils, sesquiterpenes and diterpenes, particularly phenolics, and lignans (Lin et a l, 1999; Karawya et al, 1986; Ozaki et al., 1983).
Lin et al., (1999) isolated 84 compounds from the leaves of Chamaecyparis formosensis, which comprised 18 sesquiterpenes, 40 diterpenes, 8 flavones, 7 lignans
and 11 other compounds.
Figure 1.8.5a: (+)-T erpinen-4-ol and (+)-sabinene
OH
(+)-T erpinen-4-ol (+)-Sabinene
The monoterpenes sabinene, terpinen-4-ol, a-pinene, (+)-3-carene and (+)-limonene have been reported as major constituents of Chamaecyparis essential oils (Karawya et al, 1986; Ozaki et al., 1983; Cheng and von Rudloff, 1970). Sesquiterpenes include thujopsene, P-bisabolene, cedrol and cadinenes (Andersen and Syrdal, 1970;
Cheng and von Rudloff 1970). Sesquiterpene alcohols were isolated from C. lawsoniana by McDaniel (1989) who identified t-cadinol, ô-cadinol (torreyol) and a- cadinol as having termiticidal activity. C. nootkatensis heartwood oil contains nootkatane sesquiterpenes including nootkatene, nootkatone and valencene, whereas the curcumene sesquiterpenes predominate in the leaf oil (Andersen and Syrdal,
1970).
4 7 - Figure 1.8.5b: (+)-Ô-Cadinene (-)-p-Curcumene Nootkatene
Ferruginol and similar phenolic diterpenes are major components of the
Chamaecyparis genus and of other genera from the Cupressaceae.
OH OH OH
HOOC
Figure 1.8.5c: Ferruginol Pisiferic acid Pisiferdiol
Pisiferic acid and its congeners, including pisiferol and pisiferal and other compounds such as pisiferin and pisiferdiol with an expanded B ring, have been isolated from C. pisifera (Xiao et al., 2001; Yatagai and Nakatani, 1994). Two new diterpenes, nootkastatin 1 and 2 , which showed activity against human cancer cell lines, were isolated from C. nootkatensis (Pettit et al., 2004).
- 4 8 - pH ,OH
HO. OH
OH
OH O.
Figure 1.8.5d: Nootkastatin 2 Quercetin
Flavones isolated from this genus include the antioxidant quercetin, amentoflavone, sequoiaflavone and hinokiflavone (Lin et al., 1999).
The triterpenes chamaecydin, isochamaecydin and chamaecydinol were isolated from the seeds of C obtusa by Hirose et al., (1983).
O
HO.
Figure 1.8.5e: Chamaecydin
- 4 9 - Lignons found in Chamaecyparis include hinokinin and the related lignons chamalignolide and j3-hydroxyhinokinin isolated from C obtusa (Kuo et al. 2002).
Figure 1.8.5f: Chamalignolide
OH
50 1.8.6 Pinus
There are over 100 species of the genus Pinus which is a member of the Pinaceae.
Pines are found mainly in the Northern Hemisphere, in temperate regions, or in
mountainous areas of tropical regions. The needle-like leaves are borne in fascicles
of two to five needles. The resinous cones usually ripen in the second year,
occasionally in the third year (Vidakovic, 1991). In some Pinus species, the resin
melts in high temperatures, or in situations of forest fire, which cause the cones to
open and the seeds are released.
1.8.7 Pinus nigra
Pinus nigra Arnold is also known as European Black Pine or Austrian Pine. It is indigenous to southeastern Europe and northwest Africa and will tolerate a variety of
soil types. The needles are arranged in pairs; the cones are 4 - 8 cm long, 2 - 4 cm
across, light brown and resinous when immature and do not open until the third year of growth (Vidakovic, 1991).
Figure 1.8.7a: P. nigra Figure 1.8.7b: Immature cone and shoot of P. nigra
-51 - 1.8.8 Ethnobotany of Pines
Other genera of the Pinaceae, particularly Juniperus species (Digrak et a l, 1999;
Leung and Foster, 1996) have been used in traditional medicine, however, the majority of ethnobotanical information focuses on the use of pines. This may be due, in part, to their clean, fresh aroma, which even today is used to sell cleaning products. Pines have been used by many different cultures in traditional medicine, primarily to treat wounds, rheumatism and respiratory complaints. Pine resin, bark, needles and cones have all been used. The Carrier people of British Columbia,
Canada, use pitch from the Lodgepole Pine, Pinus contorta, mixed with lard as an ointment to treat wounds and bites. Pitch is also chewed or swallowed for healing sore throats. They use the inner bark to treat colds and chew the tips of young needles as a health tonic (Ritch-Krc et al., 1996). In traditional Chinese medicine, oleoresin from P. tabulaeformis and P. massoniana is used to treat rheumatism, boils and sores (Leung and Foster, 1996). A hot water extract of P. parviflora cones is used in parts of Japan to help patients with gastric cancer (Sakagami et al., 1991). In
Turkey, indigenous species of pine, including P. nigra are used in traditional medicine. The leaves are used to remove mucus and as an antiseptic (Digrak et al.,
1999). In the UK, pine resin has been used as an antiseptic oil and was occasionally chewed to treat throat infections (Mabey, 1996).
1.8.9 Chemistry of the Pinus Genus
The chemistry of the Pinus genus is largely comprised of monoterpenes found in essential oils from various parts of the tree, the resin acids found in oleoresin, phenolic compounds, stilbenes and lignans.
5 2 - Monoterpenes which form the major constituents of Pinus essential oils are: a- and
P-pinene, limonene and myrcene (Macchioni et al., 2003), with 3-carene, P- phellandrene, a-terpinene, bomyl acetate, a-terpineol and terpinolene also present in substantial amounts (Cheng and von Rudloff, 1970).
CH;
H,C CH,
a-Pinene p-Pinene (+)-Limonene
Figure 1.8.9a: a and P-pinene and limonene
The chemistry most often associated with the Pinus genus, is that of the diterpene resin acids. The major Pinus resin acids are of the abietane, pimarane and labdane types. The most commonly found abietane acids in Pinus species are abietic acid and its isomers neoabietic acid, levopimaric acid and palustric acid and one of its breakdown products, dehydroabietic acid. Pimarane acids include pimaric acid and isopimaric acid, and the major labdane acids found in Pinus oleoresin are pinifolic acid, 4-epiimbricataloic acid and cis- and ^r<2«5 -communic acids (Phillips and
Croteau, 1999; Ucar and Fengel, 1995; Zinkel et a l, 1985).
- 5 3 - Palustric acid Pimaric acid
COOH
Figure 1.8.9b: Examples of ,CH: abietane, pimarane and labdane resin acids
tO O H
Pinifolic acid
Phenolic compounds including the stilbenes, and tannins including proanthocyanidin polymers are also chemical components of the Pinaceae. Pinosylvin is a stilbene found in the heartwood of Pinus species.
HO
Figure 1.8.9c: Pinosylvin \ / HO
- 5 4 - 1.8.10 Present Day Use of Conifer Compounds
Several studies have revealed antiviral activity for conifer extracts which supports the traditional use of conifers to treat infections. An isolate from Pinus nigra cones showed anti-HIV activity which was thought to be due to a phenylpropanoid component (Eberhardt and Young, 1996). Antiviral activity and antitumour activity has also been reported for Pinus parviflora cone extracts (Sakagami et al., 1991).
Debiaggi et al. (1988) reported that an ethanolic extract of the leaves of C. lawsoniana had antiviral activity against Herpes simplex virus type 2.
Pycnogenol® is a water extract of the bark from the French Maritime pine, Pinus maritime, produced by Horphag Research (www.pycnogenol.com). The product contains flavanoids, catechins, epicatechin, procyanidins and ferulic acid and caffeic acid and is marketed as “a natural antioxidant supporting healthy circulation”.
However, there are numerous reported biological activities for Pycnogenol®, including: antimicrobial activity against Gram-positive and Gram-negative bacteria, yeasts and fungi (Torras, et al., 2005); anti-inflammatory, antithrombotic and antioxidant activity; improvement in the lung function of asthma suffers,
(Rohdewald, 2002); antidiabetic activity (Liu, et a l, 2004); wound healing (Blazso et al., 2004). At least one study has disputed the antioxidant claims for Pycnogenol®.
Silliman et al. (2003), did not find any antioxidant effect on blood or urine samples of healthy young adults after consumption of 200 mg Pycnogenol® per day over a two week period.
The huge success story for conifers is, of course, the anticancer drug taxol isolated from the bark of the Pacific Yew tree, Taxus brevifolia. Taxol (also called
- 5 5 - paclitaxel) is prescribed to treat ovarian and breast cancer. Initially, taxol was found only in the bark of this slow growing species of yew and it would require the bark of two or three trees to treat one patient. The unusual mode of action of taxol was reported by Susan Horowitz and co-workers in 1979 (Schiff et al., 1979). Many plant anticancer compounds such as podophyllotoxin, colchicine, vinblastine and vincristine act as spindle poisons, inhibiting the formation of the spindle at mitosis.
Taxol however, enhances formation of the spindle, but then prevents its breakdown.
This new mode of action led to renewed interest in taxol, despite the shortage of material. The total synthesis of taxol was described by Nicolaou et al., in 1994, but the lengthy process was unsuitable for commercial production of the compound.
However, their research did identify an active reaction intermediate, similar to taxol, called docetaxel. This compound was not a natural product and was patented as taxotere by Rhone Poulenc Rorer. Further investigation identified the presence of similar taxoid compounds, particularly 1 0 -deacetyl baccatin 111 in the needles of the
English yew, Taxus baccata, which had the benefit of not needing to kill the tree to extract the compound. Taxol or the related compound, docetaxel can be produced via the semi-synthetic route fi*om 10-deacetyl baccatin 111. The story of taxol shows that a potent compound has been found in conifers which resulted in a beneficial drug. It emphasises the importance of taxonomy in identifying related species which, if they do not contain the compound of interest, may possess similar compounds which can then be used in the semi-synthesis of the drug. A common criticism of plant compounds is that they are complex and difficult or impossible to synthesise, however the story of taxol does demonstrate that when there is sufficient pressure or incentive this problem can be surmounted.
- 56 - 1.8.11 Conifer Oleoresin
Conifers secrete oleoresin as a defense mechanism against predators, which include bark beetles and fungal pathogens. Oleoresin, also known as “pitch”, comprises two fractions: turpentine which contains the more volatile mono- and sesquiterpenes, and the rosin or colophony fraction which is mainly formed of diterpenes, particularly resin acids (Langenheim, 2003). The turpentine fraction has a dual role; many of the monoterpenes it contains such as limonene and a- and P-pinene are toxic to beetles, and it also acts as a carrier vehicle for the more viscous diterpenes. On exposure to air the turpentine evaporates, leaving the rosin which polymerises to form a semi crystalline mass which seals the site of attack, and traps any invading beetles in the sticky resin (Phillips and Croteau, 1999).
A complex relationship exists between conifers and their predatory beetles and fungal pathogens (Kopper et al., 2005; Phillips and Croteau, 1999). Conifers are often predated by beetles of the genera Dendroctonus and Ips which bore into the bark then mate and lay their eggs in the cambial tissue. After hatching and larval pupation, the adult beetles bore their way out of the tree. The beetles act as vectors for ascomycete fungi from the Ophiostomataceae and invasion by these two pathogens results in the death of the tree. Studies by Kopper et al. (2005) determined that the monoterpenes of Pinus resinosa had activity against the beetle Ips pini and the resin acids abietic, isopimaric dehydroabietic acid were most active against the fungus Ophiostoma ips, reducing germination and mycelial growth. Some species of bark beetle have the ability to modify monoterpenes to oxygenated derivatives which they then use as sex attractants or aggregation pheromones which signal to other beetles to attack the tree (Phillips and Croteau, 1999).
- 5 7 - Only conifers from the Araucariaceae and Pinaceae produce substantial amounts of resin (Langenheim, 2003). Pinus species constitutively produce oleoresin and store it in secretory ducts or canals. Other conifers only store small amounts in resin cells
{Thuja spp.) or blisters {Abies spp.) and production is induced by insect attack
(Langenheim, 2003; Phillips and Croteau, 1999). Some conifers, such as the
Cephalotaxaceae, do not produce any resin (Langenheim, 2003).
- 5 8 - 1.9 TERPENES
All of the compounds isolated in this study were either sesquiterpenes or diterpenes and therefore a brief description of terpene biosynthesis is given below.
All terpenes are derived from the C 5 isoprene unit. They have carbon skeletons made up of C5 units, hemiterpenes (C 5), monoterpenes (Cio), sesquiterpenes (C 15), diterpenes (C 20), Isoprene triterpenes (C 30) and tetraterpenes (C 40) (Dewick, 1997).
With the exception of the hemiterpenes, the starting point for all terpene synthesis is two C5 units, isopentenyl pyrophosphate (IPP) and dimethylallyl pyrophosphate
(DMAPP).
OPPOPP
IPP DMAPP
Figure 1.9a: Isopentenyl pyrophosphate and dimethylallyl pyrophosphate
Monoterpenes are formed from the reaction between DMAPP and IPP catalysed by the enzyme prenyltransferase to yield geranyl diphosphate (GPP). The reaction is thought to start with the ionization of DMAPP to a cation, involving the loss of pyrophosphate. This is followed by attack on the cation by the double bond of IPP, giving a cationic intermediate. Then, the stereospecific loss of a proton from the neighbouring carbon to the cation occurs, resulting in the formation of a double bond and the product geranyl diphosphate (Heinrich, et a l, 2004, Dewick, 1997).
- 5 9 - OPP
DMAPP IPP [pro-R
Cationic intermediate OPP
^pro-S ^pro-R
Geranyl pyrophosphate OPP
Figure 1.9b: Biosynthesis of GPP (from Heinrich et a l, 2004)
GPP can then go on to form many different monoterpenes via different reactions (one of the most common being cyclisation), catalysed by terpene synthase enzymes.
Despite using the same precursor IPP, the synthesis of this precursor molecule of different classes of terpenoids does not occur by the same metabolic pathway, or even in the same location in the plant. The synthesis of IPP to produce
-60- sesquiterpenes and triterpenes takes place in the cytosol via the mevalonate pathway.
However, the precursor IPP for the production of the monoterpenes, diterpenes and tetraterpenes is synthesised via the pyruvate-glyceraldehyde-3-phosphate pathway in the plant plastids (Phillips and Croteau, 1999).
.IPP
Prenyltransferase
GPP Famesyl pyrophosphate (FPP)
Figure 1.9c: Formation of the sesquiterpene precursor FPP (from Dewick, 1997)
The sesquiterpenes are formed by the addition of one molecule of IPP to GPP. The product, famesyl pyrophosphate then undergoes many different reactions which produce a vast range of sesquiterpene compounds, including the germacranes, artemisinins, caryophyllene and the sesquiterpene dimer, gossypol.
OPP
Figure 1.9d: Geranyl geranyl pyrophosphate
-61 - Synthesis of the diterpenes is by the addition of IPP to FPP, again with the formation of a cationic intermediate, resulting in the end product of geranyl geranyl pyrophosphate (GGPP). GGPP can then be used in the production of numerous diterpene compounds. In conifers, the most commonly occurring diterpenes are of the abietane, pimarane or labdane class (San Feliciano and Lopez, 1991).
-62- 2.0 MATERIALS AND METHODS
2.1 PHYTOCHEMICAL METHODS
2.1.1 Conifer Material
For the initial screening of conifer species, plant material was collected from Paxton
Hill, Great Paxton, Cambridgeshire. Immature cones of Pinus nigra and Chamaecyparis lawsoniana for large scale extraction were also collected from Great Paxton. Cones of
Chamaecyparis nootkatensis and five other Pinus species were provided by Bedgebury
Pinetum, Goudhurst, Cranbrook, Kent TNI7 2SL, UK.
A voucher specimen (ECJS/OOl - ECJS/016) of each plant was placed in the herbarium at the School of Pharmacy.
2.1.2 Solvent Extraction
For the initial screening, small amounts (<50 g) of cones were chopped and ground using an IKA Werke grinder. For the Pinus cones, an electric saw had to be employed.
Sequential extraction using IL of solvents of increasing polarity: - hexane, EtOAc and
MeOH was carried out with sonication at room temperature. The extracts were filtered and IL of the same solvent was added to the conifer material for further extraction. To remove traces of water, anhydrous sodium sulphate was added to the hexane and EtOAc extracts (Houghton and Raman, 1998). Solvent was removed from the extracts under vacuum using a BÜCHI Rotavapor R-205 rotary evaporator at 40°C. A freeze dryer was used to remove water from the MeOH extracts. All extracts were stored at -20°C.
63- Large scale extraction was carried out using a Soxhlet apparatus. Solvent is held in a
round bottomed flask which sits on a heating mantle. As the solvent evaporates, the
vapour hits the condenser where it drips onto the plant material held in a thimble. The
solvent and extract will fall down into the flask where the extract gradually becomes
more concentrated. The advantage of this method is that pure solvent vapour constantly
passes through the plant material and therefore only a relatively small volume of solvent
is required, compared with cold solvent extraction whereby extraction will stop once the
solvent becomes saturated (Williamson et al., 1996). The cones were exhaustively
extracted using a series of solvents of increasing polarity: 3.5L hexane, CHCI 3, acetone
and MeOH. Extracts were dried using a rotary evaporator and, where necessary, a
freeze dryer was employed to remove remaining water. Finally the plant material was removed from the Soxhlet thimble, placed in two 1 L beakers, and water was added.
The material was sonicated for five hours, then filtered and freeze dried.
2.1.3 Thin-layer Chromatography (TLC)
Analytical thin-layer chromatography (TLC) is a method whereby compounds are
separated between a stationary phase (TLC plate) and a mobile phase (solvent).
Samples are spotted on a line 2 cm from the bottom of the plate which is then placed in a
tank with sufficient solvent to wet the plate, but not to cover the samples. As the solvent
rises up the plate, individual compounds will migrate to a point on the plate according to
their polarity. On a silica plate, polar compounds will have greater affinity for the silica
than for the non-polar solvent and will not migrate very far up the plate. Such
compounds will have small retardation factor {Rf) value (Gibbons and Gray, 1998):-
-64 Distance compound travelled from origin
= ______
Distance solvent front travelled from origin
Therefore, silica plates are used to separate non-polar compounds, whereas reverse phase Ci 8 plates, where a hydrocarbon chain is linked to the silica, are used for more polar compounds. Non-polar extracts and compounds were run on 20 x 20 cm
aluminium backed plates coated with 0.2 mm silica gel 60 F 254 (VWR). These plates
contain a fluorescent compound, which when illuminated with ultraviolet (UV) light at
254nm will give a bright background. Compounds which absorb UV light at this
wavelength will appear as dark bands against the bright background. Plates were also
examined at a wavelength of 366 nm to detect fluorescent compounds. The plate can be
developed using an appropriate spray to stain the bands. Vanillin/sulphuric acid spray
(4% vanillin in concentrated sulphuric acid) is a universal spray which will stain many
classes of compound; for example, terpenes should appear as blue or purple bands.
Alternatively, a more specific spray can be used such as Dragendorff s Reagent which
detects alkaloids. In normal phase adsorption chromatography, polar compounds may
be adsorbed onto the silica. Polar groups on the compound can interact with the polar
silanol groups on the silica plate which may result in tailing of the band. This situation
can be improved by the addition of acetic acid (1 - 2%) to the solvent system (Gibbons
and Gray, 1998). After the plate was sprayed, heat was applied to develop the bands.
2.1.4 Preparative Thin Layer Chromatography (PTLC)
One of the ftmctions of analytical TLC is to develop a suitable solvent system which can
be scaled up for larger scale preparative TLC (PTLC). Glass backed silica plates
- 65 - coated with 2 mm silica gel, F 254+366 (VWR) were used initially for PTLC, but the results were unsatisfactory, with the sample running very differently from the analytical plates.
This may be due to the larger silica particles used on PTLC plates (Gibbons and Gray,
1998). A more consistent result was achieved by using several analytical TLC plates, loading a maximum of 10 mg of sample per plate. To detect the bands, a thin strip was cut off the plate and sprayed with vanillin/sulphuric acid. This strip was then matched up with the rest of the plate so that the desired bands could be marked in pencil and scraped off the plate. The silica was placed in an elution column and washed several times with the appropriate solvent to elute and recover the sample. The sample was then dried on a rotary evaporator and under nitrogen gas.
2.1.5 Vacuum Liquid Chromatography (VLC)
Vacuum liquid chromatography (VLC) is often employed as the first fractionation step of a crude extract. A 14 x 6.5 cm glass column, fitted with a T-piece adaptor was packed with silica gel 60 (VWR), (Cig can be used for polar extracts). The vacuum was briefly applied and the silica was packed down with a glass beaker to ensure the column was tightly and evenly packed to avoid cracking (Pelletier et al., 1986). The column was washed with the first elution solvent to remove silica fines and equilibrated under vacuum. Up to 10 g of crude extract was added to an equal weight of silica and dissolved in the appropriate solvent. The solvent was removed using a rotary evaporator until the silica with the adsorbed extract formed a free flowing powder (Coll and
Bowden, 1986). This powder was added to the top of the column and 200 ml fractions were eluted under vacuum with a gradient of 10% increments, for example, going from
- 66 100% hexane to 100% EtOAc. Finally, the column was washed with EtOAc/MeOH
50:50. The fractions were spotted on a TLC plate then dried using a rotary evaporator.
2.1.6 Biotage^^ Flash Chromatography
Flash chromatography gives a finer separation than VLC and is a technique which is normally used after crude fractionation such as VLC has been carried out. Nitrogen gas is used to provide a positive pressure to the mobile phase which then moves quickly through the system. The sample (for example, an active VLC fraction) is run on analytical TLC to develop a suitable solvent system and the compound of interest should have an value of around 0.35 (Still et al., 1978), or if several compounds are required, the solvent system should be adjusted until the Rf of the least polar compound is 0.35.
The solid phase, a Biotage™ pre-packed 40M silica column 15x4 cm was used (Cig columns are also available). The column was washed with 2 x 250 ml hexane to remove silica fines, followed by equilibration with 2 x 250 ml of the elution solvent. The elution solvent can be either a gradient or an isocratic system. Approximately 500 mg of sample were dissolved in the appropriate solvent and added to 1 g silica gel 60, the solvent was then removed using a rotary evaporator until a free flowing powder was formed. This sample was packed into a small column 10 x 2.5 cm. Approximately 60 fractions were collected and spotted on TLC. Fractions with the same TLC profile were pooled and dried down using a rotary evaporator.
2.1.7 Solid Phase Extraction (SPE)
Solid phase extraction (SPE) is another adsorption technique involving separation of a sample between a mobile and a solid phase. This method can be used after crude
- 67 - fractionation by VLC to give a finer fractionation. Alternatively, SPE may be employed to clean-up a sample, for example, to retain contaminants whilst the desired compounds
are eluted or vice versa (Hostettman et al., 1998). The sample should first be run on
analytical TLC to determine a suitable solvent system. A pre-packed Strata silica column (4 X 2.5 cm) was placed on top of the vacuum manifold and equilibrated by the addition of 50 ml of the appropriate solvent. A negative pressure was applied by vacuum to force the solvent through the column. The sample (up to 500 mg), was dissolved in no more than 1 ml of solvent and was applied to the top of the colunrn. The vacuum was applied to draw the sample onto the column. The vacuum was then released and the first 50 ml aliquot of the eluent (isocratic or gradient system) was added and the vacuum reapplied. The fractions were collected in beakers placed in the system tank. Some of the solvent was removed from the fractions on a rotary evaporator before spotting the samples on a TLC plate.
2.1.8 Size Exclusion Chromatography
Size exclusion chromatography, also referred to as size inclusion chromatography, is a process whereby compounds are separated according to their size and is a useful tool to remove large contaminant molecules such as chlorophyll from a sample. Large molecules are not included in the gel matrix and will elute first, whereas the passage of
small molecules down the column will get delayed as they pass through the matrix, thereby eluting according to their size. There may also be some interactions between the compounds and the gel which will affect their elution profile (Gibbons and Gray, 1998).
Sephadex LH-20 (Amersham Biosciences) is a beaded gel manufactured by the hydroxypropylation of Sephadex G-25 (G-25 is made by cross-linking dextran with
-68- epichlorohydrin) and is suitable for use with organic solvents. 100 g of Sephadex LH-20
were added to 500 ml of CHCI 3 and left to swell overnight. The slurry was quickly poured into a glass column (54 x 4 cm) fitted with a frit and tap, and allowed to settle.
When CHCI 3 is the eluent, the Sephadex being less dense, will float on top of the
solvent, therefore the tap was opened to remove the excess CHCI 3 and draw the
Sephadex down onto the base of the column. Care has to be taken that the top of the column does not dry out. 2 0 0 mg of sample were dissolved in the minimum volume of
CHCI3 and added to the top of the column, the tap was opened to draw the sample onto the gel matrix. CHCI 3 was added drop wise from a separating funnel to avoid disturbing the top of the gel. Once all of the colour had been eluted from the colunrn, it was washed with CH 2CI2, followed by CH 2Cl2/MeOH 50:50 and finally 100% MeOH. 10 ml
fractions were collected and spotted onto TLC. Fractions with the same TLC profile were pooled and dried by rotary evaporation.
2.1.9 High Performance Liquid Chromatography (HPLC)
High Performance Liquid Chromatography (HPLC) uses colunms which are tightly packed with particles of small diameter (3-10 pm) and high pressure is needed to drive the solvent through the column resulting in good resolution (Stead, 1998). Analytical
HPLC is a versatile technique which can be used to develop a suitable solvent system for
sample scale up on preparative HPLC, or it may be employed in metabolomic studies to
achieve a profile of the compounds present in a crude mixture. The sample is injected
onto the column and the eluting solvent pumped through the system under pressure, passing through the column to the UV detector. A gradient or isocratic solvent system can be used. A Waters 600 controller with a 717 plus autosampler and 996 photodiode
-69- array detector was used. The system was controlled by Millenium 3.2 software. A
reverse phase XTerra™ MS Cig 150 x 4.6 mm colunrn (Waters) with a 5 pm pore size
was used. 30 pi samples at a concentration of 1 mg/ml, were loaded onto the column at
a flow rate of 1 ml/min. Samples were passed through a Watman 0.45 pm PVDF
syringe filter before loading. All solvents were degassed with helium.
Preparative HPLC is a technique usually employed for the final separation of
compounds from a sample which has already been purified and, ideally, contains only two or three main components. Preparative HPLC was carried out on a Waters Prep LC
4000 system with a 2487 dual X absorbance detector. A reverse phase XTerra^^ MS C 18
300 X 19 mm column (Waters) with a 10 pm pore size was used. Samples of
approximately 50 mg up to a maximum of 100 mg were dissolved in 1 ml solvent, and the system was run at 50 ml/min. Fractions (30 ml) were collected using an ISCO Foxy
200 fraction collector and a Kipp & Zonen BD40 chart recorder.
2.2 SPECTROSCOPIC METHODS
2.2.1 Nuclear Magnetic Resonance (NMR)
Nuclear Magnetic Resonance (NMR) spectroscopy is a powerful tool used to determine the structure of a compound or to give information on the type of compounds present in
an extract (Heinrich et aL, 2003).
NMR spectra were recorded on Bruker AVANCE 400 and 500 MHz spectrometers. All
samples were dissolved in deuterated chloroform and spectra were calibrated to the chloroform peak - 7.24 ppm for *H spectra and 77.0 ppm for ^^C. Spectra were processed using the X-win NMR 3.5 software package.
-70- 2.2.1.1 H NMR
The H NMR spectrum gives information on the number of protons present in a
compound and their chemical environment. It will also give an indication of the purity
of a sample. Integration of the spectrum using the area under the peaks reveals the
relative number of protons associated with each peak. The resonance of each proton is
given a chemical shift value 5 which is reported in parts per million (ppm). When the
nucleus is in a shielded (electron donating) environment, a higher field strength is
required to cause resonance resulting in low 5 values. Therefore, the ppm indicates
whether the protons are in a deshielded (high ppm) or shielded (low ppm) environment
(Kemp, 1978). The proton spectrum reveals the types of ftmctional groups likely to be present, for example, the proton of an aldehyde group gives a characteristic signal at 9.5
- 10.5 ppm and aromatic protons resonate between 6 - 8 ppm. Methyl groups in a
shielded environment give signals around 0.7 - 1.5 ppm.
Coupling constants known as J values can reveal the position of protons in respect to
each other, for example whether they are in the ortho or meta position in an aromatic
ring, or, in the case of a double bond, whether they are cis or trans to each other.
The proton spectrum can reveal considerable information about a sample and can be used as a dereplication tool, for example, to identify samples containing undesired
chemical groups or types of compounds. Where the proton spectrum indicated a pure
compound, full NMR spectra, both one- and two-dimensional were acquired to
determine the structure.
-71 2.2.1.2 ‘^C NMR
Carbon NMR uses the eomparatively rare *^C isotope as is not magnetic. Two types of carbon spectra were acquired. The broadband decoupled carbon spectrum gives a peak for each carbon atom present in a compound. The peak is a singlet, since the protons are decoupled during the experiment. The DEPT-135 (Distortionless
Enhancement by Polarisation Transfer) spectrum only shows a signal for carbons with a proton attached and facilitates identification of methine, methylene and methyl groups attached to the carbons. Comparison of the two spectra will reveal quartemary carbons, which do not show up in the DEPT experiment.
2.2.1.3 Two-dimensional spectra
The and experiments are referred to as one-dimensional techniques displaying the data along the x and y axes. A two-dimensional spectrum shows experimental data additionally on the z axis, which displays the intensity of the signal. The data may be from the same experiment, or from two separate experiments, for example proton data along the jc axis and carbon data on the y axis
The techniques which follow are all two-dimensional.
2.2.1.4 Correlation Spectroscopy (COSY)
The COSY technique provides information on proton coupling. It is described as homonuclear as it acquires only data from hydrogen nuclei. However, the data is displayed along three axes on the spectrum, and it is therefore a two-dimensional technique. The chemical shift data is plotted on the diagonal and proton coupling is
-72 shown by cross peaks which lie off the diagonal axis, indicating protons which are only two-to-four bonds from each other.
2.2.1.5 Heteronuclear Single Quantum Coherence (HSQC) Spectroscopy
This technique is heteronuclear as it uses data acquired from both proton and carbon nuclei. An HSQC spectrum reveals which protons are attached to which carbons.
Quartemary carbons will not give a signal since they do not have any protons attached.
2.2.1.6 Heteronuclear Multibond Coherence (HMBC) Spectroscopy
The HMBC technique provides long-range correlation information on protons and carbons which are separated by two or three bonds. From the information given in the
HMBC spectmm, the stmcture of the compound can begin to be assembled. This is done by assembly of partial fragments of the compound.
2.2.1.7 Nuclear Overhauser Effect Spectroscopy (NOESY)
The NOESY spectrum is another homonuclear technique involving proton data only. It gives information on the spatial proximity of protons, therefore protons which are not close together in terms of bond distance, but are close together in space may give a
signal. The NOESY technique can therefore facilitate in assigning the relative
stereochemistry of a compound.
-73 2.2.2 Infra-red Spectroscopy and Polarimetry
An infra-red spectrum of each compound was acquired using a Nicolet 360 FT-IR
spectrophotometer. Optical rotations were acquired on a Bellingham and Stanley ADP
2 0 0 polarimeter.
2.2.3 Gas Chromatography-Mass Spectrometry (GC-MS)
All the isolated compounds were analysed by gas chromatography combined with mass spectrometry (GC-MS). These experiments were carried out by Neale Wareham at
Stiefel Laboratories, Maidenhead, using an Agilent 6890 GC coupled to an Agilent 5973 mass selective detector. An HP-5ms capillary column of 30 m length with a diameter of
250 pm was used with a non-polar stationary phase of 5% phenylmethylsiloxane and a
film thickness of 0.25 pm. Samples were introduced into the system using the split
injection technique with a split ratio of between 5:1 and 10:1 and an injector temperature of 250°C. Helium was used as the carrier gas at an average linear velocity of 50 cm/sec.
The initial oven temperature was 50°C and the temperature was increased after 5 minutes at a rate of 5°C/min to a maximum of 300°C. The MS was run in electron
impact mode and a chromatogram was generated for each sample from the total ion count. A search for the main peaks was performed using the NIST98 library.
Gas liquid chromatography employs a gas mobile phase with a liquid stationary phase held on a solid support. The sample is injected into the system, vapourised at high temperature and carried in the gas to the column. Samples will be separated according to their volatility and eluted in order of their boiling points (Robards et aL, 1994).
74- 2.3 BIOLOGICAL METHODS
2.3.1 Minimum Inhibitory Concentration (MIC) Assay againstS. aureus
Minimum Inhibitory Concentration (MIC) assays were performed in 96 well microtitre plates (Nunc) using a broth dilution method. Mueller-Hinton broth (MHB; Oxoid) was
adjusted to contain 20 mg/L Ca^^ and 10 mg/L Mg^^. NaCl (2%) was added to MHB for
assays against MRSA strains in accordance with NCCLS guidelines. To assay for
activity against MRSA strains, oxacillin was used in place of methicillin, also in
accordance with NCCLS guidelines. Overnight cultures of each S. aureus strain were made up in 0.9% saline to an inoculum density of 1 x 10^ colony forming units (cfu) by comparison with a McFarland standard. They were then diluted with saline to 5 x 10^ cfu. All antibiotics were obtained from Sigma Chemical Co. Tetracycline and oxacillin were dissolved directly in MHB, whereas norfloxacin and erythromycin were made up
first in dimethyl sulphoxide (DMSG). Samples were made up in DMSG, then diluted in
MHB to give the desired starting concentration in well 1 of the microtitre plate (usually
512 pg/ml). 125 pi of MHB were dispensed into wells 1-11, then 125 pi of the test
compound or the appropriate antibiotic were dispensed into well 1 and serially diluted
across the plate, omitting well 11 which was the growth control. The final volume was
dispensed into well 12, which being free of MHB or inoculum, served as the sterile
control. Finally, 125 pi of the bacterial inoculum were added to wells 1 - II and the plate was incubated at 37°C for 18 hours. A DMSG control (3.125%) was included and
all MICs were determined in duplicate.
The following day, the plate was developed using a methanolic solution (5 mg/ml) of 3-
[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT; Lancaster).
-75- Metabolically active bacteria reduce the yellow MTT to a dark blue formazan and it is thought the reduction may be carried out by mitochondrial dehydrogenase enzymes
(Mosmann, 1983). The MIC was determined as the lowest concentration at which no
growth was seen.
2.3.2 Modulation Assay
Modulation assays were carried out in 96 well microtitre plates. To test for potentiation
of antibiotic activity, compounds or extracts were dissolved in DMSO, then in MHB to
give the desired concentration. The MHB was then used as in the MIC assay. As a
control for potentiation of tetracycline, norfloxacin or erythromycin activity against
effluxing strains of S. aureus^ the efflux pump inhibitor reserpine (Sigma) was
incorporated into MHB to give a concentration of 20 pg/ml. For potentiation of
oxacillin activity against EMRSA-15 and -16, epicatechin gallate (EG; Sigma) was
incorporated into MHB to give a concentration of 10 and 4 pg/ml respectively. A
DMSO (3.125%) control was included.
Some compounds were additionally assayed to ascertain whether the presence of a known modulator of antibiotic activity enhanced the activity of the compound, which would be shown by a reduction in MIC for that compound. Compounds were assayed in
combination with reserpine (20 jig/ml) against the effluxing strains XU212 and SA-
1199B. Against EMRSA-15 and -16, epicatechin gallate was included at 10 and 4 pg/ml
respectively.
-76- 2.3.3 S, aureus Strains used in MIC and Modulation Assays
The S. aureus strains used in the MIC and modulation assays and their MICs against four standard antibiotics are summarised in table 2 .1 .
The standard strain ATCC 25923 was supplied by E. Udo. Five clinically révélant strains of S. aureus were used in the assays to encompass assays against effluxing, MDR and MRSA strains. Three effluxing strains were used in the assays:
XU212 is a Kuwaiti hospital isolate, obtained from E. Udo (Gibbons and Udo, 2000).
This strain is multidrug-resistant, it possesses the TetK pump which effluxes tetracycline, it is an MRSA strain and is highly resistant to erythromycin (MIC 4,096 pg/ml);
SA1199B is a clinical bloodstream isolate provided by G. Kaatz (Kaatz et al., 1993). It over-expresses the Nor A MDR efflux pump which removes fluoroquinolones such as norfloxacin, antiseptics and quaternary ammonium compounds (QACs) from the cell;
RN4220 has the MsrA pump which effluxes macrolides including erythromycin and was provided by J. Cove (Ross et al, 1989).
Two epidemic MRSA strains were also used:
EMRSA-15 (Richardson and Reith, 1993) and EMRSA-16 (Cox et al., 1995) were obtained from P. Stapleton. These were the major strains found in MRSA bacteraemia isolates in a study from 26 UK hospitals (Johnson et a l, 2001). Both strains are also highly resistant to erythromycin, with EMRSA-16 also resistant to norfloxacin.
-77 Table 2.1 MIC values (^g/ml and jiiM) for standard antibiotics tested against S, aureus strains used in MIC and modulation assays
S. aureus tetracycline norfloxacin erythromycin oxacillin strain R >16pg/ml R >16pg/ml R >8pg/ml R >4pg/ml ATCC 25923 0.25 1 0.25 0.125 XU212 128 16 4,096 256 SA1199B 0.25 32 0.25 0.25 RN4220 0.25 2 128 0.25 EMRSA-15 0.125 0.5 2,048 32 EMRSA-16 0.125 128 4,096 512
S. aureus tetracycline norfloxacin erythromycin oxacillin strain R>33.3pM R^O .lpM R>10.9pM R ^ .lu M ATCC 25923 0.5 3.1 0.3 0.3 XU212 266.2 50.1 5,581.1 580.0 SA1199B 0.5 1 0 0 .2 0.3 0 .6 RN4220 0.5 6.3 174.4 0 .6 EMRSA-15 0.3 1 .6 2,790.6 72.5 EMRSA-16 0.3 400.9 5,581.1 1,159.9
R = MIC value which indicates antibiotic resistance
2.3.4 Ethidium Efflux Assay
The ethidium efflux assay was carried out by Professor Glenn Kaatz of the John D.
Dingeil Department of Veterans Affairs Medical Center, Detroit, Michigan, USA. The
assay was carried out using S. aureus strain SA-1199B which overexpresses the Nor A
efflux pump. Ethidium bromide (EtBr) which fluoresces when bound to DNA, is a
substrate for this pump and its efflux from the cell and the resulting reduction in
fluorescence can be measured using a spectrofluorophotometer. A compound which is
an inhibitor of Nor A should prevent, or at least reduce the efflux of EtBr and therefore a
strong fluorescent signal should be seen when compared with a sample which has EtBr but no inhibitor. The ethidium efflux experiment was carried out as previously described (Kaatz et al., 2000)
78 3.0 RESULTS
3.1 PRELIMINARY SCREENING
Crude extracts of six species of conifer cones were obtained by cold solvent extraction as described in section 2.1.2. The extracts were screened for activity against the standard S. aureus strain ATCC 25923 and the multidrug-resistant clinical isolate
XU212. All extracts were assayed at a starting concentration of 512 pg/ml.
The MIC assay results (Table 3.1) show that for all species, the hexane extracts were the most active, although for some species the EtOAc extracts were as active. The MeOH extracts were the least active and in Cupressaceae species, they were inactive at 512 pg/ml. The Pinus nigra MeOH extract did show some activity but was not as active as the hexane extracts. Cephalotaxus harringtonia cones differ from those of the Pinaceae and Cupressaceae. The cone, described as a drupe-like fruit (Vidakovic, 1991), is the size of an olive with a green fleshy outer coat which was separated from the white inner seed. The hexane extraction yielded so little extract that it was not tested. Only the
CHCI3 extract of the outer coat showed any anti-staphylococcal activity. It is possible that the seeds of C harringtonia are more protected since they are not exposed on the scales of a cone and therefore they may not need protective compounds, or it may be that such compounds are not constitutively produced but synthesized under conditions of microbial invasion. Of course, the fact that an extract is not active against S. aureus does not preclude the possibility that it is active against other bacteria or conifer microbes.
-79 In many cases, the crude extracts were slightly more active (2-4 fold) against the MDR
XU212 than against the standard strain. This finding has been noted in other studies
(Lechner et a l, 2004; Tegos et a l, 2002) in which some isolated compounds have greater activity against MDR strains than against sensitive strains.
The hexane extract of Chamaecyparis lawsoniana cones showed the highest anti- staphylococcal activity and larger scale bioassay-guided fractionation was carried out.
Pinus nigra cones were also selected for further investigation since tlie crude hexane and
EtOAc extracts had moderate anti-staphylococcal activity and also due to the wealth of information on the use of pines in traditional medicine (Digrak et al., 1999; Leung and
Foster, 1996; Ritch-Krc etal, 1996).
-80 Table 3.1: MICs (|ig/ml) of cone extracts from Coniferae species against S. aureus
Family Species Extract ATCC 25923 XU212 (TetK)
Cephalotaxaceae Cephalotaxus CHCI3 256 256-512 harringtonia seed coat Acetone >512 >512 MeOH >512 >512
Cephalotaxus CHCI3 >512 >512 harringtonia seed Acetone >512 >512 MeOH >512 >512
Cupressaceae Chamaecyparis Hexane 16 8 lawsoniana
EtOAc 32 8 MeOH >512 >512 Cupressaceae Cupressus Hexane 64 16 abrahams EtOAc 128 32 MeOH >512 >512 Cupressaceae Cupressus Hexane 32 16 arizonica EtOAc 128 64 MeOH >512 >512 Pinaceae Abies koreana Hexane 64 n/t EtOAc 128 n/t MeOH 512 >512 Pinaceae Pinus nigra Hexane 128 64 EtOAc 128 64 MeOH 256 256 Tetracycline 0.25 128 n/t = not tested
-81 3.2 EXTRACTION OF COMPOUNDS FROM CHAMAECYPARIS LAWSONIANA
Soxhlet extraction was performed on 400 g of immature cones of C lawsoniana.
Sequential extraction started with hexane, then CHCI 3 followed by acetone and finally
MeOH. MIC assays on the crude extracts confirmed that the hexane extract was the most active and further bioassay-guided fractionation was carried out. VLC on 8 g of hexane extract yielded 12 fractions. Elution started with 100% hexane going to 100%
EtOAc in 10% increments, finishing with an EtOAc/MeOH 50:50 wash.
Table 3.2.1: MICs (pg/ml) of VLC extracts against S. aureus
VLC fraction ATCC 25923 XU212 (TetK) 2 (337 mg) >512 >512 3 (391 mg) 2 2 4 (2 .6 g) 2 2 5 ( 1.1 g) 16 8 6 (1.4 g) 32 16 7 (832 mg) 32 32 8 (444 mg) 128 128 9 (235 mg) 256 512 10 (58 mg) 512 512
Solid phase extraction (SPE) was performed on 450 mg of VLC fraction 4. Using a normal phase silica colunm, elution commenced with 1 0 0 % petroleum spirit followed by a gradient of 5% increments of diethyl ether to 70:30 petroleum spirit/EtzO. The EtiO increments were then increased to 40%, 60% and 80%, finally eluting with 100% EtiO.
Most of the sample eluted in fractions 2 and 3, which were both active in the MIC assay at 2 - 4 jLig/ml. These two fractions were used in PTLC (petroleum spirit/toluene 80:20
1 development; petroleum spirit/ EtzO 2 developments) to isolate ferruginol.
82- Figure 3.2a: SPE of VLC fraction 4 (TLC plate sprayed with vanillin/H 2S0 4 )
ferruginol
trans-communie acid
Fractions 7-11 were run on PTLC, solvent system hexane/ Et20/E t0 Ac 80:15:5, two developments, which yielded /ra«5-communic acid.
VLC fraction 6 (200 mg) was loaded onto a Sephadex column and eluted with CHCI3, followed by dichloromethane, 50:50 dichloromethane/MeOH, finishing with 100%
MeOH. Fractions with the identical TLC profiles were pooled.
-83- Figure 3.2b: TLC of pooled Sephadex column fractions (sprayed with vanillin/H 2S0 4 )
/
/ D \
A = oplopanonyl acetate B = torulosal C = formosanoxide D = 5-epipisiferol E = pisiferol Pooled fractions were run on PTLC as follows:
Fractions 12 - 17, hexane/ EtiO/EtOAc 80:15:5, three developments, gave oplopanonyl acetate and torulosal.
Fractions 36 - 42, hexane/ EtzO/EtOAc 80:15:5, four developments, yielded formosanoxide, epipisiferol and pisiferol.
Table 3.2.2: Yield of compounds isolated from 400 g immature cones of C.lawsoniana
Compound Yield (mg) Ferruginol 43.1 Oplopanonyl acetate 15.1 5-Epipisiferol 14.9 Torulosal 10.3 4P-Hydroxygermacra-1 ( 10),5-diene 6.2 Pisiferol 6.1 rra«5-Communic acid 5.5 Formosanoxide 4.2
-84- Soxhlet Extraction 400 g immature cones
Figure 3.2c: Bioassay guided fractionation of Hexane C lawsoniana extract I VLC
Fraction Fraction Fraction 4 6
Sephadex PTLC SPE LH -20
4p- Hydroxygennacra- Fraction^ { Fraction^ Fraction^ Fraction^ l(10),5-diene ' 2-3 ^ k 6-11 y . 12-16; ' 36-42/1
I 1
PTLC PTLCPTLC PTLC
Oplopanonyl Pisiferol /ra«5-Communic Ferruginol acetate 5-Epipisiferol acid Torulosal Formosanoxide
-85- 3.3 PHYTOCHEMICAL CHARACTERISATION OF COMPOUNDS FROM CHAMAECYPARIS LA WSONIANA
3.3.1 CL-001 (Ferruginol)
CL-001 was isolated as a pale yellow oil. The GC-MS data indicated a molecular formula of C20H30O with the NIST98 library giving a 72% probability that the compound was ferruginol. The ^^C-NMR data were in accordance with the published literature for ferruginol (Wenkertet a l, 1976). Six signals between 110 and 150 ppm suggested the presence of an aromatic ring. Comparison with the DEPT-135 spectrum
(Fig. 3.3.1b) revealed that four of these resonances were absent, indicating a tetra- substituted aromatic ring.
The *H-NMR spectrum showed a septet at 5 3.09 which integrated for one proton, characteristic of the methine proton of an isopropyl group. Two methyl doublets at 5
1.21 and 5 1.22 completed the isopropyl group. Signals at 5 4.49 in the proton spectrum and at 5 150.7 in the carbon spectrum were indicative of a phenolic hydroxyl group which, in the HMQC experiment gave no signal, but in the HMBC spectrum showed proton to carbon correlations. The above data strongly suggested that the compound was a phenolic abietane diterpene.
2 10 g A B .3 5 7^
Figure 3.3.1a: CL-001
86- i i n w w /
-"1-^ T 'I ...... 160 150 140 130 120 110 100 90 80 60 50 40 30 20 10 ppm
11 14
150 140 130 120 110 100 90 80 70 60 50 40 30 ppm
Figure 3.3.1b: '^C carbon and DEPT-135 spectra of CL-001
-87- [^vD(NœincNcr»coc^fOif)ü{Nr*t^{r» m'TOCoonvûcnm 938B8S8ag2F:)e|C:g!Çqi8 S S3 œ œ t' r- r- r- r- ^ G G l2S!2a8S[%$$ ■^nmmm(N(N(NMCS(NO) H-11 OH H-14 H-15 H-7 I 1 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 CNC^ xomvoro^ooro^ir)a\^(NComr^o.H*ïoaiLr)tH^OtHr^in»HXtHm^x)CNfsr^roroat\£>tHmvûr^rHcj> m M cNocriXLn'3*mxœr'in^xxfnmvox^tH.-troxir)^^mtHi-toxcoin^rnfNiTHiHox[^r^fNi<3*'^fNX XX [^[^XXXXXXXXXX^^^^XrOXXXCNOJfNOJfNfNlCNfNCN'H.-l.H.HtHrHiHt-i.HOOOOXXXX r4 (—I (Hi—4.—itHt—(i—).Hi— ItHTHtHtHtHt— (.HtH i—itHi—ItHiHi—ItHiHtHiHi—(iHtHrHWrHrHrHiHOOOO H-16 H-20 H-18 H-19 /17 H-2b H-2a H-5 H-3b H-6b H-6a H-la 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1.0 0.9 ppm Figure 3.3.1c: *H-NMR spectra of CL-001 Table 3.3.1a: (500 MHz) and (125 MHz) NMR spectral data for CL-001 in CDCI3 Position ‘H (Jin Hz) ^^C (Wenkertet al. 1976) 1 1.38 (13.0, 3.5) 38.9 38.7 2.15 (12.5) 2 1.58 r (3.5) 19.3 19.2 1.71 J r (14.0, 3.5) 3 1.18 J(3.5) 41.7 41.6 1.45 m 4 33.4 33.3 5 1.30 (12.5, 2.0) 50.4 50.3 6 1.65 m 19.2 19.2 1.84 m 7 2.77 m 29.8 29.6 2.83 brdd{lA 1.5) 8 127.3 126.1 9 148.7 147.9 1 0 37.5 37.3 11 6.61 s 1 1 1 .0 1 1 0 .6 12 150.7 151.1 13 131.4 131.7 14 6.81 s 126.6 126.2 15 3.09 sep 26.8 26.6 16 1.21 J (7.0) 2 2 .6 * 22.4* 17 1.22 J (7.0) 22.7* 2 2 .6 * 18 0.92 5 33.3 33.1 19 0.90 5 2 1 .6 21.4 2 0 1.155 24.8 24.6 OH(12) 4.49 5 * interchangeable values Correlations seen in the HMBC experiment further substantiated that CL-001 was ferruginol. From the HMBC spectrum, the isopropyl methine H-15 exhibited a V correlation to C-13 and V correlations to C-12 and C-14, with the methyl groups of the isopropyl (C-16, C-17) also showing V correlations to C-13, which placed the isopropyl group at position 13 on the C ring. The H-11 methine exhibited a V correlation to C-12 carrying the phenolic OH group and V correlations to the aromatic carbons C -8 and C- 13. The H-14 methine showed V correlations to the isopropyl C-15 and to C-7 and also -89- to the aromatic C-9 and C-12. The two aromatic protons (H-11, H-14) were both singlets and they did not couple indicating they were in a para position. Table 3.3.1b: Long-range ^H— connectivities for CL-001 detected in an HMBC experiment Position Correlated C-atom HMBC (H^C) VV 1 C-2 c-3, C-5 2 C-3 3 C-2 C-5 5 C-4, C-10 C-7, C-9, C-18, C-19 6 C-5, C-7 C-8 , C-10 7 C-6 , C-8 C-5, C-9 11 C-12 C-8 , C-13 12 OH C-12 C-11, C-13 14 C-7, C-9, C-12, C-15 15 C-13, C-16, C-17 C-12, C-14 16/17 C-15 C-13 18 C-4 C-3, C-5, C-19 19 C-4 C-3, C-5, C-18 2 0 C-10 C-1, C-9 On the A ring, the two methyl groups attached to C-18 and C-19 displayed V connectivities to the quaternary C-4 and V connectivities to C-3 and C-5. The H-5 methine in turn exhibited V correlations to C-4 and C-10 and a V correlation to C-7. The position of carbon 6 was confirmed by the H 2 -6 methylene which showed V correlations to C-5 and C-7 and V correlations to C -8 and C-10. The H 3-2 O methyl displayed a V correlation to C-10 at the A/B ring junction and V correlations to C-1 on the A ring and the aromatic C-11 on the C ring. ^H—>^H connectivities observed in the COSY spectrum identified couplings between methylene groups on C-1, C-2 and C-3 on the A ring, with the C-1 and C-3 methylenes both coupling to the C-2 methylene. These observations, together with the similarity of the ^^C data to that reported by Wenkert et al. (1976), confirmed that CL-001 was ferruginol (8,1 l,13-abietatrien-12-ol). -9 0 - The relative stereochemistry of ferruginol was determined by examination of the NOES Y spectrum and the observed coupling constants for H-5 in the 'H spectrum. The H-5 methine had a large axial coupling {J= 12.5 Hz) to the axial H -6 proton and a small coupling {J= 2.0 Hz) to the H -6 equatorial proton. This confirmed H-5 as occupying an axial position. From the NOESY spectrum, a through space interaction was observed between the H 3-2 O and H 3-I9 methyls indicating that they were on the same face of the molecule. This relative stereochemistry was assigned a (P) orientation as reported in the literature. The methine proton H-5, assigned as («), showed a NOE coiTelation to the CH 3 -I 8 methyl group which would also have an (ot) orientation. These data determined the A/B ring junction as trans. The NOESY spectrum also revealed 1-3 interactions between the axial proton H- 6 a which showed correlations to the H 3-I 9 and H 3-2 O methyls, and the equatorial protonH- 6 b which correlated to the methyl group 3H-I 8 . 20 ppm 0.9-i □ 0 0 0 1.0^ 20 1.24 1.34 1.4^ 1.54 1 . 6 - 1.7r 1.84 1.9- 2.04 2.14 2.2 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1.0 0.9 ppm Figure 3.3.Id: NOESY spectrum for CL-001 -91 - ppm 0 . 6 - 0 .8 - 1. 0 - 1 . 2 - 1.4- 1 . 6 - 1. 8 - oP M 3b/2a 2 . 0 - Q □ 2 . 2 - la-/2a+b 2.4- 2 . 6 - 2 .8 - 3.0- 3.2 3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 ppm C-12 OH ppm 2 0 - O 6 0 - 8 0- 1 0 0 - C-11 1 2 0 - C-13 140- C-12 160 - 7.5 7.0 6.56.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm Figure 3.3. le: COSY and HMBC spectra of CL-001 -9 2 - 3.3.2 CL-002 (Pisiferol) and CL-003 (5-Epipisiferol) CL-002 and CL-003 were isolated as yellow oils. For both compounds, the EI-MS data gave a molecular ion at m/z 302. The NMR data for the two compounds were similar. However, since the compounds migrated to different points on the TLC plate, and stained different colours after the plate was sprayed with vanillin/H 2S0 4 (purple for CL- 002; pale blue for CL-003), it was evident that the compounds were not identical. The ^^C spectra for CL-002 showed that the molecule contained 20 carbon atoms and the proton spectrum revealed a phenolic hydroxyl group at 5 4.90 and signals at Ô 3.13, 5 1.22 and 8 1.20 were indicative of an isopropyl group. In the IR spectrum, a broad absorption at 3354 cm'^ was attributed to a phenolic hydroxyl group. These data suggested that the compound was an abietane diterpene with strong similarities to ferruginol observed in the carbon spectrum. The difference of 16 Da in molecular weight between CL-002 and ferruginol suggested the presence of an additional oxygen group. The extra oxygen was assigned at C-20 as an oxymethylene moiety observed in the proton (8 3.65, 3.92) and carbon (8 64.1) spectra. This was confirmed by HMBC couplings of H 2-2 O to C-1 and C-9 and of the H-5 methine to C-20. OH OH Figure 3.3.2a: CL-002 Figure 3.3.2b: CL-003 -9 3 - Full HMBC analysis showed that CL-002 was very similar to ferruginol, but with an oxymethylene group in place of a methyl at C-20. The proton data were also in accordance with the data published by Yatagai et al (1978), for pisiferol and CL-002 was therefore determined as pisiferol (8,11,13-abietatriene- 12,20-diol). Due to the similarity of the data between compounds CL-002 and CL-003, and from HMBC analysis which showed that the compounds were very similar, it was determined that CL-003 was an epimer of pisiferol. Table 3.3.2a Long-range ^H^^^C connectivities for CL-003 detected in an HMBC experiment Position______Correlated C-atom HMBC (H—»C) VV 1 C-2, CIO C-3, C-5 2 C-1, C-3 C-10 3 C-4 C-1, C-5 5 C-4, C-6 C-20 6 C-5, C-7 C-10 7 C-6 , C-8 C-5, C-9, C-14 11 C-9, C-12 C-13, C-20 12 OH C-12 C-11,C-13 14 C-8 , C-13 C-7, C-9, C-12, C-15 15 C-13, C-16, C-17 C-14 16/17 C-15 C-13 18 C-4 C-3, C-5, C-19 19 C-4 C-3, C-5, C-18 2 0 C-1, C-10 C-5, C-9, C-11 Son et al. (2005) reported a new abietane diterpene which they erroneously named 20- hydroxyferruginol. Their data are included in the following two tables of *^C and ^H NMR data for CL-002 and CL-003, as it can be seen that “20-hydroxyferruginol” is in fact an epimer of pisiferol and that the data match those for CL-003. 94- Table 3.3.2b: (125 MHz) NMR spectral data for CL-002 and CL-003 in CDCI 3 Position '^C CL-002 "C CL-003 ‘^C 2 0 -hydroxyferruginol (pisiferol) (5-epipisiferol) (Son et a i, 2005) 1 32.8 41.8 41.6 2 18.9 18.7 18.7 3 41.7 42.4 42.4 4 33.5 34.4 34.3 5 50.1 58.0 58.0 6 18.5 24.3 24.4 7 28.7 35.4 35.3 8 128.5 133.0 132.7 9 141.7 135.5 135.0 1 0 42.5 71.6 72.2 11 112.9 118.8 118.7 12 150.3 151.6 152.1 13 132.6 133.4 133.3 14 127.5 126.6 126.6 15 26.8 26.6 26.5 16 22.5* 22.5* 22.9 17 2 2 .6 * 2 2 .8 * 2 2 .6 18 33.3 21.7 32.3 19 22.3 32.2 21.7 2 0 64.1 51.0 51.1 * interchangeable values In the spectrum for pisiferol (CL-002), the magnitude of the coupling for the H-5 methine to the axial H -6 proton (J = 12.5) revealed that proton H-5 occupied an axial position. The relative stereochemistry of the molecule was assigned by correlations seen in the NOESY spectrum (Fig. 3.3.2c). A correlation between the Hi-20 methylene protons and the H 3-I 9 methyl group established that they were on the same face of the molecule and were assigned an axial (|3) orientation as described in the literature (Yatagai et aL, 1978). Further NOE signals between proton H-5 and the CH 3-I 8 methyl and between H-5 and proton H- 6 b and the axial proton of the H-7 methylene suggested an alpha orientation and a trans A/B ring junction. 95 Table 3.3.2c: (500 MHz) NMR spectral data (/in Hz) for CL-002 and CL-003 in CDCI3 Position CL-002 CL-003 20-hydroxyferruginol (pisiferol) (5-epipisiferol) (Son et a i, 2005) 1 1.17 m 1.52 m 1.61 m 2.55/(13.0) 1.84 m 1.94 m 2 1.67 m 1.44 m 1.47 m 1.87 m 3 1.25/(4.0) 1.24 m 1.30 m 1.51 m 1.41/(2.5) 1.47 m 4 quaternary 5 1.47//(12.5, 2.5) 2.64//(12.0, 2.5) 1.41 m 6 1.76 m 1.26 m 1.21 m 1.84 m 1.99 m 2.06 m 7 2.85 m 2.66 m 2.70 m 2.73 m 2.78 m 8 quaternary 9 quaternary 10 quaternary 11 6.66 s 6.66 5 6.74 5 12 (OH) 4.90 5 5.99 5 13 quaternary 14 6.89 5 6.90 5 6.91 5 15 3.13 g 3 M q 3.22 sep 16 1.22*/(2.0) 1.21*/(7.0) 1.22/ 17 1.20*/(2.0) 1.19*/(7.0) 1.21 / 18 0.93 5 0.915 0.95 5 19 0.905 0 . 8 8 5 0.92 5 20 3.65/(11.0) 2.55/(14.0) 2.61 / 3.92 f (10.0) 3.00/(14.0) 3.05/ * interchangeable values For CL-003, proton H-5 also had a large coupling (/ = 12.0) to the H-6 axial proton. The NOESY data indicated a cis A/B ring junction for the molecule. This was deduced by through space interactions between the H-20b methylene proton and the H-5 proton which in turn had a correlation to the H 3-I9 methyl indicating that these protons were all on the same face of the molecule. These protons were assigned an axial (P) configuration and CL-003 was described as 5-epipisiferol, an epimer of pisiferol. -9 6 - OH OH CL-002 CL-003 Figure 3.3.2c: Selected NOEs for CL-002 and CL-003 The proton and carbon NMR data were very similar between the two compounds, however, a number of differences were apparent. In the spectra, the H 2-2 O methylene showed a difference in resonance of approximately 5 1.0, resonating further downfield in pisiferol compared with 5-epipisiferol. From the data (Table 3.3.2c), it can be seen that in 5-epipisiferol, the H 2-2 O methylene group resonates as a set of doublets (6 2.55, 3.00, J = 14.0 Hz), indicating that the two protons couple to each other but not to the hydroxyl group. For pisiferol, these protons resonate significantly further downfield as a doublet and an apparent triplet (63.65, 3.92, J = 10.5). The signal appeared to be a triplet but, on a higher field instrument it may resonate as a double doublet resulting from one of the methylene protons coupling to the hydroxyl group as well as to its geminal proton. The hydroxyl may be preferentially held in the 71 electrons of the phenol, accounting for the difference in splitting and shift observed for the H 2-2 O protons. A correlation detected in the COSY spectrum confirms that the non-phenolic hydroxyl is seen by H-20b. The cis A/B ring junction may place the H 2-2 O methylene in a more shielded environment than a trans ring. In the cis configuration, the methylene may be closer to the n cloud of the aromatic C ring which could have a shielding effect. -9 7 - In 5-epipisiferol, the cis A/B ring junction with an axial (P) methine proton at position H-5 could affect the shape of the molecule compared with the trans ring junction found in pisiferol and ferruginol where the C-5 methine is (a). From the data for the two pisiferol molecules, the greatest difference in resonance was seen for carbons C-1, C-5, C-10 and C-20 (Table 3.3.2b) which would possibly be most affected by the change in A/B ring configuration. C-10 in particular resonates at 5 71.6 in 5-epipisiferol with a cis ring, but is more shielded in the trans ring of pisiferol resonating further upfield at 5 42.5. From a ball and stick model, the molecule appears to be more planar with a trans A/B ring junction compared with a cis configuration and this may affect the environment of C-10 with respect to the C-20 hydroxyl or the aromatic ring. A more deshielding environment could cause the carbon to resonate at a higher ô value. Conversely, carbon C-20 in pisiferol resonates downfield of C-20 in 5-epipisiferol at 5 64.1 and 6 51.0 respectively. In this instance, the carbon is more shielded in the cis configuration compared with the trans. The data for 5-epipisiferol are very close to that recently reported by Son et aL, (2005). This group reported the compound as 20-hydroxyferruginol and assigned a trans configuration for the A/B ring junction even though the carbon resonances at C-5, C-10 and C-20 are not consistent with those reported for ferruginol but are comparable with those seen here for a cis configuration. The group also determined the relative stereochemistry on the basis of NOE correlations between the Hi-20 and H-11 protons and between H-5 and H%-2. However, it is difficult to see how these NOEs would facilitate the assignment of the A/B ring junction stereochemistry. This compound is therefore reported here for the first time as 5-epipisiferol. -9 8 - 20b 20a 7 ppm A k______A JWL M 1. 0 - 1.5- oO' 2.0 2.5- 3.0- 3.5- 4.0- 'n ' ^ ' I ' ' ' ' n- ' ' ' I ' I—'—'—' ' I 4.0 3.5 3.0 2.5 2.0 1.5 20b 20a ppm 18 0 “® 1 . 0 - 1.5- 2 . 0 - 2.5- 3.0- 3.5 3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 ppm Figure 3.3.2d: NOESY spectra for CL-002 (top) and CL-003 -9 9 - c^^(Na^if)rom(Nr4aDCDr"ir)mr4\DminiAv«g'mmoor~^LALAmoo\ciinuii CDo^o^oD''^'.Or4Mr-iaococDaDaoaoif)inr-iooooaococo[-'r''r^r~r'r''''Ovovovoi H-12 OH 20b 20a 7 . 0 6 . 5 6.0 5 . 5 5 . 0 4 . 5 4 . 0 3 . 5 3 . 0 2 . 5 2.0 1 .5 1.0 ppm 8 g / fel IV£>{NO H-12 OH 20b 20a JuuJ Li_ 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm Figure 3.3.2e: *H spectra for CL-002 (top) and CL-003 - 100- 3.3.3 CL-004 (Formosanoxide) CL-004 was isolated as an amorphous white solid. The GC-MS data gave a molecular ion at m/z 300 and solved for a molecular formula of C20H28 O2. The and NMR data were similar to pisiferol and 5-epipisiferol, suggesting a further abietane diterpene. A signal at m/z 285 in the mass spectrum (Fig. 3.3.3d) indicated the loss of a methyl group, with further peaks at m/z 270 and 255 indicating the loss of two and three methyls respectively. Signals in the proton spectrum at 5 3.17 (H-15 sep IH) and ô 1.24 (H-16/17 dd 6H) and aromatic resonances in the carbon spectrum hirther supported that the compound was an abietane diterpene. Carbon signals at 5 70.3 and ô 67.7 indicated the presence of an oxygen atom attached to these carbons. The molecular weight of 300 Da for CL-004 was 2 Da less than for the pisiferol epimers which could be accounted for by the loss of two hydrogen atoms if the compound contained an ether bridge rather than an oxymethylene group at C-20. OH Figure 3.3.3a: CL-004 - 101 - Table 3.3.3a: (125 MHz) NMR spectral data for CL-004 in CDCI 3 Position ‘H (Jin Hz) ‘^C Formosanoxide *^C data Hsu et al. 1 1 .2 1 J(1.5) 2 & 8 2 & 8 1.72 m 2 1.65 m 18.8 18.8 1 .8 8 m 3 1.55 m 41.2 41.2 4 33.8 33.7 5 1,19 ddd (11.5, 6.0, 2.0) 42.9 42.9 6 1.58 m 30.1 30.2 2.05 m 7 4.79 dd(3.5, 2.0) 70.3 70.4 8 131.9 131.6 9 145.3 145.2 1 0 37.5 37.5 11 6.57 s 106.9 106.9 12 OH 4.67 s 152.2 152.4 13 131.2 131.3 14 7.02 s 1 2 1 .0 120.9 15 3.17 sep 26.8 26.8 16 1.23 d{\.5) 22.7 22.7 17 1.25 d{\.5) 22.7 22.7 18 0.82 s 32.8 32.7 19 1.15 s 20.9 2 0 .8 2 0 2.82 dd(S.5, 2 .0 ) 67.7 67.7 4.31 J(8.5) On the basis that the compound was related to pisiferol, examination of the spectrum revealed that the methylene protons at C-20 resonated as a doublet (H-20b, 64.31) and a double doublet (H-20a, 62.82). The 20a proton had a large coupling to 20b (J= 8.5 Hz) and a small coupling to H-5 {J= 2.0 Hz) with the protons exhibiting a W coupling in the COSY spectrum. Inspection of the HMBC spectrum revealed a correlation between both H 2-2 O protons and C-7 suggesting an ether bridge between carbons 7 and 20. H-5 was assigned an axial position, determined by one if its couplings to H -6 ( / = 11.0 Hz) observed in the proton spectrum. The NOESY spectrum showed through space 102 correlations between H-20b and the H-19 methyl group and between the axial H-5 methine and the H-18 methyl. These groups were assigned a (P) and an (a) orientation respectively, as indicated in the literature (Hsu et al. 1995). Formosanoxide was isolated from Chamaecyparis formosensis by Hsu et al. (1995) and the ^H and NMR data for CL-004 were consistent with that reported, therefore CL-004 was determined as formosanoxide (7 p,20-epoxyabieta-8,11,13-trien-12-ol). Table 3.3.3b: Long-range ^H—>^^C connectivities for CL-004 detected in an HMBC experiment Position______Correlated C-atom HMBC (H—»C) 1 C-2 c-3, C-5 2 C-3 3 C-2 C-5 5 C-4, C-10 C-7, C-9, C-18, C-19 6 C-5, C-7 C-8 , C-10 7 C-6 , C-8 C-5, C-9 11 C-12 C-8 , C-13 12 OH C-12 C-11,C-13 14 C-7, C-9, C-12, C-15 15 C-13,C-16, C-17 C-12, C-14 16/17 C-15 C-13 18 C-4 C-3, C-5, C-19 19 C-4 C-3, C-5, C-18 2 0 C-10 C-1,C-11 103 ■ 7 .240 3 1.00_ > - ■ 7 .019 ft W 1.00 _ m 6.572 W w cr 13 O S C/D o 2 3 o ’-b O 1.02 K) r (To 8 g 4^ I »■ h ' 18 19 20b 20a ppm j LJL 1.0 1.5 2 . 0 2.5 3.0 3 . 5 4.0 19/20b 4.5 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm Figure 3.3.3c: NOESY spectrum of CL-004 - 10 5 - TIC; 2705T03 0 (i) Figure 3.3.3.d: GC-MS data for CL-004 (i) GC trace (ii) Total Ion Count (iii) Mass Spectrum 5500000 5000000: 4500000 10 00 45 00 (ii) Signal : TIC peakR.T. first max last PK peak c o r r . c o r r . % of min scan scan scan TY h eig h t area % max. t o t a l 1 3.043 153 184 233 BB 3 55014 889436 0.13% 0.124% 2 41.566 7493 7542 7586 BB 4 68276 2863790 0.41% 0.400% 3 42.179 7618 7659 7705 BB 4 61406 2928280 0.41% 0.409% 4 42.886 7741 7794 7911 BB 2 16156121 706810385 100.00% 98 .668! 5 44.069 8005 8020 8070 BB 4 53017 2863342 0.41% 0.400% 100 270 (iii) 50 199 157 300 227 255 115128 145 188 55 69 ^,1? i 239 i„ 285 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 - 106- 3.3.4 CL-005 (4p-Hydroxygermacra-l(10),5-diene) CL-005 was isolated as a clear oil. The GC-MS data showed a molecular ion at 222 m/z. A peak at 204 m/z for a loss of 18 Da was characteristic for the loss of water. Another peak at 179 m/z arising from the loss of a 43 Da fragment could result from the loss of an isopropyl group. A further peak at 161 m/z could account for the loss of an isopropyl group and water. The carbon spectrum showed the molecule contained 15 carbons; four signals were in the aromatic/oleflnic region of the spectrum and one carbon resonated at Ô 73.1 characteristic of a carbon attached to an hydroxyl group. The proton spectrum revealed the likely presence of an isopropyl group, with two doublets at 5 0.77 and ô 0.81, each integrating for three protons, and a multiplet at 5 1.42 which integrated for one proton. This data solved for a molecular formula of C 15H26O and suggested the molecule might be a sesquiterpene. 13 DH Figure 3.3.4a: CL-005 107- Table 3.3.4a: (500 MHz) and (125 MHz) NMR spectral data for CL-005 in CDCI3 Position ‘H (Jin Hz) ''C ^^C (Cornwell et a\. 2001) 1 4.93 b rd {\\.5 ) 128.9 128.8 2 1.93 w 23.7 23.6 2.48 m 3 1.55 J (4.0) 39.6 39.6 1.62 J r (14.5, 4.0) 4 73.1 73.0 5 5.23 J (15.5) 140.1 140.0 6 5.16 (16.0, 9.5) 125.7 125.7 7 1.99 m 52.8 52.8 8 1.37 m 25.9 25.9 9 2.23 m 41.3 41.2 1 0 132.6 132.5 11 1.42 m 33.0 33.0 12 0.81 J (7.0) 2 0 .6 2 0 .6 13 0.77 J (6.5) 19.0 18.9 14 1.52/(1.0, 2.5) 16.7 16.7 15 1.175 30.7 30.7 The NMR data were in close agreement with those reported for a sesquiterpene germacradienol by Cornwell et al. (2001). This was confirmed by examination of the ^H and HMBC spectra. The ^H experiment revealed a large coupling (J= 15.5 Hz) between the methine protons H-5 and H -6 indicating a trans double bond for the vinyl group. H- 6 also had a coupling (7= 9.5) to the proton attached to C-7. Correlations were observed in the COSY spectrum between the two methyl groups CH 3-I2 and CH 3-I 3 to the H-11 methine, which in turn had a correlation to C-7. These data suggested that C-11, C-12 and C-13 formed an isopropyl group which was attached to C-7 and was confirmed by correlations in the HMBC experiment which revealed V connectivities between the CH3-I 2 and CH 3-I 3 methyls and the H-11 methine and V connectivities to C-7. The H- 11 proton also showed a V correlation to C-7 and a V correlation to C- 6 . A correlation was observed in the COSY spectrum between the methines H -6 and H-7 with a V - 108- correlation in the HMBC spectrum from the H-7 proton to C -6 and a V correlation to C- 5. The H -6 proton showed V connectivities to C-5 and C-7 and a V connectivity to C-4. Table 3.3.4b: Long-range ^H—>^^C connectivities for CL-005 detected in an HMBC experiment Position Correlated C-atom HMBC (H—>C) V 1 - C-9, C-14 2 C-1 C-4 3 -- 5 C-4, C-6 C-7 6 C-5, C-7 C-4 7 C-6 C-5 8 C-7, C-9 C- 6 , C-10 9 C-8 , C-10 C-1, C-7, C-14 11 C-7, C-12, C-13 C-6 12 C-7, C-11, C-13 - 13 C-7, C-11, C-12 - 14 C-10 C-1, C-9 15 C-4 C-3, C-5 The quaternary C-4 had been identified from its resonance at 5 73.1 in the *^C spectrum as having an attached hydroxyl group. A V connectivity from the CHs-15 methyl to C- 4, observed in the HMBC spectrum also determined that this methyl group was attached to C-4. V correlations from the CH 3-I 5 methyl to C-3 and C-5 were also observed. The H%-2 methylene had a V connectivity to C-1 and a V connectivity to C-4, with correlations to H-1 and H 2-3 in the COSY spectrum. This also determined the position of C-1 together with HMBC V correlations to C-9 and C-14. A V connectivity from the CH3-I 4 methyl to C-10 and V connectivities to C-1 and C-9 confirmed that this methyl group was attached to C-10. A correlation was observed in the COSY spectrum from the H-9 protons to the H 2-8 methylene which in turn had a correlation to H-7. Further connectivities were seen in the HMBC experiment with V correlations from the H 2-9 109- protons to C-1, C-7 and C-14, and from the H 2-8 methylene to C -6 and C-10. The data were in accordance with that reported in the literature (Cornwell et al., 2001; San Feliciano et al., 1995) and CL-005 was assigned as 4(3-hydroxygermacra-l(10),5-diene (P-germacrenol). The hydroxyl group was assigned a (3 orientation, as reported by Bohlmann et al. (1984) who determined that the a and (3 4-epimers could be distinguished by the chemical shifts of the H-5 and H -6 protons. Where H-5 resonates slightly downfield of H- 6 , the hydroxyl has a (3 orientation, as seen here (H-5 = 5 5.23, H -6 = 55.16). However when H-5 is found slightly upheld of H- 6 , the hydroxyl group occupies an a position. C7>COiH(NO>OrOO cy>inrooo»-H 12/ 11/8 ILi 7/2a 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm 0> O (O 0*1 IN P O) V h- N m Figure 3.3.4b: *H spectrum of CL-005 -110- 3.3.5 CL-006 (/#*fl/is-Communic acid) Compound CL-006 was isolated as a very pale yellow oil. The GC-MS data gave a molecular ion at 302 m/z with peaks at 287, 273 and 257 m/z which could correspond to the loss of a methyl, a methyl and a methylene group, and a carboxylic acid group respectively. The *^C spectrum solved for 20 carbons, indicating that CL-006 was likely to be a diterpene. Examination of the IR spectrum showed peaks at 1693, 1736 and 2931 cm '\ characteristic of signals for the hydroxyl and carbonyl of a carboxylic acid group. The ^^C spectrum had seven signals in the aromatic/olefinic region of the spectrum, one signal at 5 182.5 was characteristic for a carboxyl group and from the DEPT-135 experiment it was seen that two of the other carbons were methylene groups and that the compound therefore contained two exomethylene groups. From the GC-MS and NMR data, CL-006 solved for a molecular formula of C20H30O2. Figure 3.3.5a: CL-006 18^ GOGH 111 - Table 3.3.5a: Long-range ^H— connectivities for CL-006 detected in an HMBC experiment Position______Correlated C-atom HMBC (H—>C) V 1 C-3 2 -- 3 C-1, C-18 5 C-4, C-6 C-18, C-19, C-20 6 C-7 C-8, C-10 7 C-8 C-9, C-17 9 C-8, C-10, C-11 C-17, C-20 11 C-9, C-12 C-8 12 C-9, C-14, C-16 14 C-12, C-16 15 C-14 C-13 16 C-13 C-12, C-14 17 C-7, C-9 18 C-4 C-5, C-19 20 C-10 C-5, C-9 The 'h and ^^C NMR data were similar to the labdane diterpene communie acid, previously isolated from Chamaecyparis obtusa by Yamamoto et al. (1997) and from Calocedrus formosana by Fang et al. (1989). The data from the HMBC experiment also supported this observation. The H 2-I 5 methylene exhibited a V correlation to C-14 and a V correlation to the quaternary C-13 and the H-14 methine had V correlations to C-12 and the methyl group carbon C-16. The H-12 methine had V connectivities to C-9, C-14 and C-16. These data suggested that the C-14, C-15 vinyl group formed the end of a hydrocarbon chain. From the ^H spectrum, the J coupling constants for the vinyl group protons identified the H-15b proton at 5.03 as occupying a trans position across the double bond (/= 17.5 Hz) in relation to the H-14 methine. Proton H-15a with a J value of 11.0 Hz occupied the cis position. The H 2- 1 1 methylene showed V correlations to C- 9 and C-12, with a V correlation to the oleflnic C-8. The H-9 methine had V connectivities to C-8, C-10 and C-11, with V connectivities to the oleflnic C-17 and the - 112- C-20 methyl. These data for H-9 suggested a cyclohexane ring system. The 2-7 H methylene displayed a V correlation to -8C and V correlations to C-9 and C-17. The H- 5 methine showed V connectivities to C-4 and C-6 and V connectivities to the CH3-I 8 and C H 3-2O methyl groups and also to the carboxyl group C-19. The CH3-2 O m ethyl exhibited a V correlation to C-10 and V correlations to C-5 and C-9. Comparison of these data with those of other diterpenes isolated in this study, suggested that C-5 and C- 10 were at the junction of two cyclohexane rings. On the A ring, the CH3-I 8 m ethyl showed a V connectivity to the quaternary C-4 and V connectivities to C-5 and C-19. The position of C-3 was determined by V correlations from the H-3 methylene to C-1 and C-18. The H-1 methylene, in turn displayed a V correlation to C-3. This left only one position for the signal at 5 19.1 which was assigned to C-2. All the diterpenes isolated in this study with A/B cyclohexane rings have a value of approximately6 19 for C -2. Comparison of the data with that reported in the literature (Yamamoto et al., 1997; Fang et al., 1989), with the differences explained below, determined CL-006 as trans- communic acid. There were differences in the assignment of some of the carbon atoms between the three sets of NMR data and those differences have been highlighted in red in table 3.3.5a. All the NMR spectra were recorded in CDCI3, therefore the use of different solvents cannot account for the differences and it can be seen that the data are extremely close in terms of chemical shift, but the actual assignment of the position of some of the carbons differs. Yamamoto et al. used a 400 MHz spectrometer and Fang et al. did not state the instrument used. However, the data reported in this study were obtained on a 500 MHz - 113- spectrometer which may have facilitated the interpretation of the data, compared with the instruments available to the other two groups. Table 3.3.5b: (500 MHz) and (125 MHz) NMR spectral data for CL-006 in CDCI3 Position *H(yin Hz) '"C ^ C (Fang et al. Yamamoto 1989) et al. 1997) 1 1.05 m 38.0 39.2 37.9 2.16 m 2 1.52 m 19.1 19.8 19.9 3 1.14 (13.5, 4.0) 39.2 38.4 39.2 1.86 m 4 44.1 44.1 40.4 5 1.34 (12.0, 2.5) 56.2 56.3 56.3 6 1.96 m 25.8 25.8 23.3 7 1.91 m 38.5 37.7 38.5 2.39 m 8 147.9 147.8 147.9 9 1.76 s 56.4 56.2 56.4 10 40.3 40.3 44.2 11 2.13 m 23.3 23.3 25.8 2.36 m 12 5.39 t (6.5) 133.9 141.5 133.9 13 133.4 133.4 133.5 14 6.31 dd{\1.5, 11.0) 141.6 133.8 141.6 15 4.86 6/(11.0) 109.9 109.9 109.9 5.03 6/(17.5) 16 1.73 5 11.8 11.8 12.8 17 4.45 5 107.7 107.6 107.7 4.82 5 18 1.23 5 29.0 29.0 29.0 19 182.5 183.7 184.1 20 0.64 5 12.8 12.8 11.9 Fang et al. transposed the signals for the vinyl group carbons C-12 and C-14, but their position was unambiguously assigned from the splitting pattern observed in the ^H spectrum. The H-14 methine resonated as a double doublet due to the two neighbouring protons across the double bond, attached to C-15. The coupling constants revealed a cis - 114- coupling to proton H-15a (J = 11.0 Hz) and a trans coupling to H-15b {J = 17.5 Hz). The H-12 methine, however, resonated as a triplet as it was seen by two equivalent protons, the H 2-II methylene, giving a / value of 6.5 Hz. The COSY experiment also revealed strong correlations between the H-12 and H-11 protons and another correlation between the H-12 and H-9 methines. There were also differences in the assignment of C-1, C-3 and C-7. The positions of these carbons were determined by correlations observed in the COSY and HMBC spectra. The H:-17 methylene had an allylic coupling to the H 2 -7 methylene in the COSY spectrum and a V correlation was observed in the HMBC spectrum. The position of C-7 was further reinforced by a V correlation from the H 2 -6 methylene to C-7. It was difficult to determine which carbon signal should be assigned to C-1 and C-3. The CH 3- 20 methyl would usually show a ^/correlation to C-1 which assists in the positioning of this carbon, but no correlation was detected in the HMBC experiment. There were no correlations in the COSY spectrum for the H 2-I or H2-3 methylenes, but from the HMBC spectrum, a V correlation from H2-3 to C-18 determined the position of the signal of C-3. The H 2-3 methylene had a V connectivity to C-1 and the H 2-I methylene had a V correlation to C-3. This led to the unambiguous assignment of these signals. Yamamoto et al. transposed the positions of the quaternary carbons C-4 and C-10. However, V correlations from the H-9 methine and CH 3-2 O methyl established the position of the signal due to C-10 at the A/B ring junction and V connectivities from the H-5 methine and CH 3-I 8 determined the position of the signal of C-4. They also transposed the positions of C -6 and C-11, but strong correlations between the H 2-II methylene and H-12 methine in the COSY and a V correlation to C-12 in the HMBC firmly established the position of the signal of C-11. Yamamoto et al. also switched the - 115- positions of the CH 3-I 6 and CH 3-2 O methyl groups. The CH 3-I 6 methyl is connected to C-13 which forms a double bond with C-12 and they may have placed the downfield carbon 5 12.8 in position C-16 considering it to be deshielded due to the double bond. It is interesting that, although C-20 is downfield at 6 12.8 in comparison to C-16 at 5 11.8, the CH 3-2 O protons resonated upheld at 8 0.64 in comparison to the CH 3-I 6 protons which were at 8 1.73. The CH 3-I 6 methyl had connectivities in the HMBC spectrum to C-12, C-13 and C-14 which determined the position of this methyl group on the molecule. Vand V correlations to C-5, C-9 and C-10 also established the assignment of the CH 3-2 O methyl group. The H-5 methine had a large axial coupling {J= 12.0 Hz) to the axial proton H -6 of the C-6 methylene. In the NOESY experiment a NOE was observed between the H-5 methine and CH 3-I 8 methyl and these two groups were assigned an a orientation as reported in the literature (Yamamoto et al. 1997; Fang et al. 1989). A NOE interaction was detected between the H-5 and H-9 methines; this would give the H-9 proton an axial (a) orientation, making the relative stereochemistry of the hydrocarbon chain attached to C-9 equatorial and p. A through space interaction between the CH 3-2 O methyl and CH 2- 11 methylene indicated that they were on the same face of the molecule and suggested that the methyl group occupies a position which is axial and p, giving a trans A/B ring junction. - 116 (g ft w w 1/1 cr roi_ g 2!ôF~ 1 2 3 o’-+1 0 r -J 1 Os oo 3.3.6 CL-007 (Torulosal) CL-007 was isolated as a yellow oil. The GC-MS data indicated a molecular ion at 286 m/z. The spectrum solved for 20 carbons with an oxygen-bearing quaternary carbon at 6 73.6 and a signal at ô 205.8 characteristic of an aldehyde. This was confirmed by examination of the IR spectrum which revealed a peak at 1715 cm'^ for an aldehyde group and a broad beak at 3400 cm'^ for the hydroxyl group. The spectrum also showed a downfield signal at 5 9.72 characteristic of an aldehyde proton. These data, together with further observation of the and DEPT-135 spectra, solved for a molecular formula of C20H32O2 with a molecular weight of 304 Da. This was 18 Da greater than the ion seen in the MS spectrum, but the loss of water could account for the difference in mass. Two signals at 5 107.4 and ô 111.8 for olefmic carbons in the spectrum were determined from the DEPT-135 experiment as two exomethylene groups. OH Figure 3.3.6a: CL-007 H - 118- Table 3.3.6a: (500 MHz) and (125 MHz) NMR spectral data for CL-007 in CDCI3 Position ‘H (yin Hz) ^'C *'C (^uetal. 1994) 1 1.83 6rrf(13.0) 34.4 34.4 1.97 m 2 1.54 m 19.3 19.2 3 1.02 m 38.5 38.4 2.11 b rd (13.5) 4 48.7 48.6 5 1.43 (13.5, 3.0) 56.1 55.8 6 1.68 m 24.1 24.0 2.03 m 7 1.97 m 38.5 38.4 2.44 m 8 147.5 147.5 9 1.56 m 55.8 55.9 10 40.3 40.2 11 1.55 m 18.0 17.9 12 1.29 Û? (8.5) 41.2 41.3 13 73.6 73.6 14 5.88 (17.5, 11.0) 145.0 145.0 15 5.04 (11.0, 1.0) 111.8 111.7 5.18 (17.5, 1.0) 16 1.25 5 28.1 28.1 17 4.50 (0.5) 107.4 107.3 4.85 (1.5) 18 0.99 5 24.4 24.3 19 9.12 d (1.5) 205.8 205.7 20 0.54 5 13.5 13.5 Assuming a labdane skeleton for CL-007, it was possible to show connectivities supportive of this diterpene class. From observation of the coupling constants for the vinyl protons (C-14, C-15) in the ^H spectrum, the geminal protons on C-15 had a very small coupling {J = 1.0 Hz), with proton H-15a having a cis coupling to the C-14 methine ( /= 11.0 Hz). Proton H-15b was identified as occupying a trans position across the double bond by its large coupling constant of J = 17.5 Hz. In the HMBC experiment, the H 2-I 5 vinyl group methylene protons showed a V correlation to C-14 - 119- and V correlation to the quaternary C-13 which carried the hydroxyl group. The C-14 methine also showed a V correlation to C-13 and it was therefore determined that C-15 was likely to be at the end of a hydrocarbon chain. 15a 17a/b 10 9 8 7 6 5 4 3 2 1 ppm IS Figure 3.3.6b: spectrum of CL-007 The CH 3-I 6 methyl group showed a V correlation to C-13 and V correlations to C-12 and C-14. The H 2-I 2 methylene displayed V connectivities to C-11 and C-13 and V connectivities to C-9 and C-14. The H-9 methine had V correlations to C-11 and the quaternary C-10 and also to C -8 which formed a vinyl group with C-17. The signal of C-9 also had V correlations to the C-17 and C12 signals and to that of the methyl group CH3-2 O. The H 2-I 7 methylene had a V correlation to C -8 and a V correlation to C-7. The H 2-7 methylene, in turn showed a V connectivity to C -8 and V connectivities to C- 17 and C-5. The position of C -6 was determined by a V correlation from the H 2-6 - 120- methylene to C-5 and a Vcorrelation to C- 8 . The H-5 methine had a Vcorrelation to C- 4 and V correlations to the methyl groups CH 3-I 8 and CH 3-2 O and the C-19 aldehyde group. This completed the B ring and placed C-5 and C-10 at the A/B ring junction. Table 3.3.6b: Long-range ^H—>^^C connectivities for CL-007 detected in an HMBC experiment Position______Correlated C-atom HMBC (H ^C) V 1 - c-5 2 - - 3 C-4 C-5 5 C-4 C-18, C-19, C-20 6 C-5 C-8 7 C-8 C-5, C-17 9 C- 8 , C-10, C-11 C-12, C-17, C-20 11 C-9, C-12 12 C-11, C-13 C-9, C-14 14 C-13 15 C-14 C-13 16 C-13 C-12, C-14 17 C-8 C-7 18 C-4 C-3, C-5, C-19 19 C-4 C-3 2 0 C-10 C-1, C-5, C-9 The CH 3-I 8 methyl and the H-19 aldehyde proton both displayed V connectivities to the quaternary C-4 and V connectivities to C-3. The CH 3-I 8 methyl also had a V correlation to C-5. The position of C-3 was determined from a V correlation to C-4 and a V correlation to C-5. The H 2-2 methylene did not exhibit any correlations in the HMBC or COSY spectra, but the H 2-I methylene did have a V correlation to C-5 and the CH 3-2 O methyl had a V correlation to C-1 which confirmed the position of C-1 which left only one position for C-2 on the A ring. The CH 3-2 O methyl also had a V connectivity to C-10 and V connectivities to C-1 and C-9 which confirmed its position attached to the quaternary C-10 of the ring junction. The data for CL-007 were - 121 - consistent with that reported in the literature (Su et al., 1994; Enzell, 1961) and CL-007 was therefore assigned as torulosal ( 13-hydroxy-8( 17), 14-labdadien-19-al). The stereochemistry of torulosal was determined from examination of the and NOESY spectra. A large coupling constant for the methine proton H-5 {J= 13.5 Hz) to the axial H -6 proton indicated an axial (a) orientation for H-5 as reported in the literature. The NOESY spectrum revealed a through space interaction between the H-5 methine and the CH 3-I 8 methyl which would also have an a orientation. The C-19 aldehyde group displayed NOEs to the methyl groups CH 3-I 8 and CH 3-2 O and the ring junction was therefore determined as trans, with the CH 3-2 O methyl and C-19 aldehyde occupying a (p) axial orientation. The CH 3-2 O methyl also exhibited a NOE to the H 2- 11 and H 2-I 2 methylenes. ppm 20 18 1 0 - 10 9 8 7 6 5 4 3 2 1 ppm Figure 3.3.6c: NOESY spectrum of CL-007 - 1 2 2 - 3.3.7 CL-008 (Oplopanonyl acetate) CL-008 was isolated as an opaque film. The GC-MS data showed a very small peak at 265 m/z, with larger peaks at 220 and 205 m/z. The spectrum solved for 17 carbons with a signal at 5 211.2 being characteristic for a ketonic carbon and another at ô 170.5 characteristic for the carbonyl carbon of an ester. A further signal for a quaternary carbon at 5 85.0 suggested an oxygen bound to a carbon atom. From the IR spectrum, peaks at 1724 and 1700 cm'^ were indicative of carbonyl groups. Examination of the ^H, and DEPT-135 spectra suggested a molecular formula for CL-008 of C 17H27O3 (280 Da). The MS peak at 265 m/z could result from the loss of a methyl group, with a peak at 220 m/z accounting for the loss of acetic acid and another at 205 m/z resulting from the loss of acetic acid plus a methyl group. These data suggested the compound could be a sesquiterpene containing an acetyl group. O 15 I I 17 O C CH Figure 3.3.7a: CL-008 -1 2 3 - Table 3.3.7a: *H (500 MHz) and (125 MHz) NMR spectral data for CL-008 in CDCI3 Position 'H (Jin Hz) 13c ^^C (De Bruyn et al., 1990) 1 1.40 m 25.6 27.0 1.76 m 2 1.56 m 28.4 29.9 1.91 m 3 2.63 m 55.5 56.8 4 1 .8 6 m 45.8 47.4 5 1.09 m 49.2 50.7 6 1.07 m 2 2 .6 24.0 1.58 m 7 1.58 m 37.0 38.5 2.54 m 8 85.0 8 6 .6 9 1.77 m 54.5 56.1 1 0 2 1 1 .2 214.6 11 2.17 5 29.5 30.0 12 1.43 m 29.5 30.9 13 0.87 J (7.0) 21.9 22.5 14 0.66 d (6.5) 15.6 16.3 15 1.45 5 17.0 17.5 16 170.5 172.7 17 1.92 5 2 2 .6 2 2 .8 The data showed strong similarity to those for the sesquiterpene oplopanonyl acetate, isolated by De Bruyn et al. (1990). This was confirmed by correlations seen in the COSY and HMBC experiments. ^H to *H correlations noted in the COSY spectrum and V connectivities in the HMBC experiment confirmed that the CH 3-I 3 and CH 3-I4 methyls together with the H-12 methine formed an isopropyl group. A V correlation from the H-12 methine to C-5 determined that the isopropyl group was attached to C-5. The H-5 methine had V correlations to C-4 and C- 6 . The H 2-6 methylene displayed V correlations to C-5 and C-7, with a V correlation to C-4. The H 2-7 methylene showed V connectivities to the quaternary C -8 and to C -6 and V connectivities to C-5 and to the CH3-I 5 methyl group attached to C- 8 . From its chemical shift value at ô 85.0, C -8 was - 124- determined to be part of an ester group. The CH3-I 7 methyl had only one correlation (^J) to C-16, which suggested that CH3-I 7 and the carbonyl C-16 formed an ester with the oxygen attached to C-14, the whole group forming an acetate group Table 3.3.7b: Long-range connectivities for CL-008 detected in an HMBC experim ent P osition______Correlated C-atom HMBC (H—»C) VV 1 C-2, C-9 C-4 2 C-1 C -4, C -10 3 C-2, C-4, C-10 C-5 4 C-5, C-9 C-1, C-8 5 C -4, C -6 , C -12 C -9 6 C-5, C-7 C-4 7 C-6 , C -8 C -5, C -15 9 C-1, C-4, C -8 C -3, C -5 11 C -10 C-3 12 C-15, C-13,C-14 C -4 13 C-12 C-5, C-14 14 C -12 C -5, C -13 15 C -8 C -7, C -9 17 C -16 T he C H 3-I 5 methyl exhibited a V correlation to -8C and V correlations to C-7 and C-9. The H-9 methine showed V connectivities to C-1, C-4 and C-8 with V connectivities to C-3 and C-5. The H-4 methine had V correlations to C-5 and C-9 with V correlations to C-1 and C - 8 . The H-3 methine displayed V connectivities to C-2 and C-4 and to C-10, which from its chemical shift value of 8 211.2 was identified as the oxygen bearing carbon of a ketone functional group. The CH3-II methyl had a V correlation to C-10 and a V correlation to C-3, with C-10 and C-11 forming an acetyl group. Position 2 of the A ring was established by a V connectivity from the H2-2 methylene to C-1 and V connectivities to C-4 and C-10. The H-1 methylene showed V correlations to C-2 and - 125- C-9 and a V correlation to C-4, De Bruynet al. (1990) isolated oplopanonyl acetate from Chamaecyparis pisifera and, due to the similarity of the data, CL-008 was identified as the sesquiterpene oplopanonyl acetate. uiLnMC) 2/4/12 2.7 2.6 2.5 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1.0 0.9 0.8 0.7 ppm § P Isl B Figure 3.3.7b: 'H spectrum of CL-008 It was not possible to postulate the relative stereochemistry for this compound as the NOES Y experiment did not reveal any useful NOE interactions. However, due to the similarity between the NMR data, the relative stereochemistry for oplopanonyl acetate was assigned as reported by De Bruynet al. (1990) with the H-4 methine and CH 3-I 5 methyl having a (p) orientation. - 126- 3.4 EXTRACTION OF COMPOUNDS FROM CHAMAECYPARIS NOOTKATENSIS A supply of immature cones ofChamaecyparis nootkatensis was obtained from Bedgebury Pinetum. Soxhlet extraction was carried out on 500 g of cones as described in the Methods section 2.1.2. Solvents were used in the order of hexane, CHCI3 , acetone and M eOH. The Soxhlet extraction yielded the following amounts of crude extracts: 14.6 g hexane, 3.5 g CHCI3, 18 g acetone, 11.2 g MeOH. The cold water extraction yielded 1.4 g. VLC was carried out on 1 0 g of the hexane extract, starting elution with 1 0 0 % hexane and going in a gradient of 10% increments to 100% EtOAc. Finally, the column was washed with EtOAc/MeOH 50 : 50 (fraction 12). VLC was carried out on 2 g of the CHCI3 extract, commencing elution with 100% CHCI3 with a gradient of 10% increments to 100% EtOAc. However, after fraction 6 , the coloured band of sample ceased to move down the silica column and fractions 7, 9 and 11 did not yield any sample, with fractions8 and 10 only producing 10 mg and 4.3 mg respectively. After eluting with 100% EtOAc, the solvent was changed to Etc Ac/acetone 50 : 50, followed by 100% acetone and finishing with acetone/MeOH 50:50 which yielded 211 mg (fraction 14). MIC assays showed that the hexane VLC fractions 4-12 were active, unlike C. lawsoniana where the activity reached a peak at fractions 3 and 4, then tailed off with each fraction. In C. nootkatensis, the activity continued into the CHCI3 VLC fractions. Fractions 5 and 7-13 were not assayed due to the lack of sample. The final fraction (14), eluted with acetone/MeOH was inactive. - 127- Table 3.4.1: MICs (jug/ml) of C. nootkatensis VLC extracts against S. aureus VLC fraction ATCC 25923 XU212 (TetK) Hexane 3 (516 mg) >512 >512 4 (130 mg) 8 4 5 (2.7 g) 2 2 6 ( 1 .6 g) 4 2 7 (1.4 g) 4 2 8 (L 9 g ) 4 2 (1.3 g) 8 4 10 (490 mg) 8 4 11(152 mg) 8 8 12 (94 mg) 32 16 CHCI3 2 (292 mg) 8 8 3 (852 mg) 4 2 4 (193 mg) 16 16 6 (217 mg) 64 32 14(211 mg) >512 >512 H PLC was carried out on the most active CHCI3 V LC fraction (3). Analytical HPLC identified acetonitrile/HiO 70:30 as a suitable solvent system and this was used for preparative HPLC. Preparative HPLC was carried out on 65 mg of CHCI 3 VLC fraction 3 and repeated several times to generate a sufficient quantity of each fraction. The sample was dissolved in acetonitrile/H 2 0 70:30 and elution was started with the same solvent and run for 30 minutes. The acetonitrile concentration was then increased in a gradient up to 100% over 5 minutes and held for 2 minutes. Detection was at 254 nm. - 128- 0 40C 8 00 800 1 0 00 1200 1400 16 00 1800 20 00 22 00 Tim e Figure 3.4.1: HPLC trace of CHCI3 VLC fraction 3 Peak 1 - Sugiol Peak 2 - 7a-Hydroxytotarol Peak 3 - insufficient sample for analysis Peak 4 - Ferruginol Peak 5 - Sempervirol Peak 6 - Totarol Table 3.4.2: Yield of compounds isolated from 500 g immature cones of C.nootkatensis Compound Yield (mg) Totarol 3& 0 7a-Hydroxytotarol 11.2 Sempervirol 4.5 Ferruginol 3.1 Sugiol 1.4 - 129- 3.5 PHYTOCHEMICAL CHARACTERISATION OF COMPOUNDS FROM CHAMAECYPARIS NOOTKA TENSIS 3.5.1 CN-001 (Totarol) CN-001 was isolated as a colourless oil. The GC-MS data gave a molecular ion at 286 m/z. The ^^C-NMR spectrum indicated the presence of 20 carbon atoms, six of which resonated in the aromatic region of the spectrum. The spectrum showed strong similarities to ferruginol, with a characteristic multiplet at 5 3.28 which integrated for one proton and a triplet (two superimposed doublets) at 5 1.32 which integrated for six protons, suggesting an isopropyl group. A singlet at 5 4.38 in the proton spectrum and at 5 151.9 in the carbon spectrum and which had no signal in the HMQC experiment, was indicative of a phenolic hydroxyl. However, there were two doublets in the aromatic region of the proton spectrum at ô 6.98 and 5 6.49 which each integrated for one proton. This indicated that these two aromatic protons must be in close proximity as they split each other’s signal giving rise to a doublet. These protons therefore occupied a different position on the aromatic ring to ferruginol where the two aromatic protons did not couple as they were para to each other and resonated as singlets. .OH Figure 3.5.1a: CN-001 •16 1 130 Table 3.5.1a: (500 MHz) and (125 MHz) NMR spectral data for CN-001 in CDCI3 Position 'H (Jin Hz) u c '"C (NishidaeraZ. 1977) 1 1.30 (4.0) 39.6 39.6 2.21 b rd (\2.5) 2 1.58 m 19.5 19.5 1.71 (14.0, 3.5) 3 1.20 (13.5, 4.0) 41.6 41.6 1.45 m 4 33.3 33.2 5 1.25 (12.5, 2.0) 49.6 49.6 6 1.64 m 19.4 19.4 1.89 (13.5, 8.0) 7 2.74 m 28.7 2 & 8 2.92 Jc/(17.0, 6.5) 8 134.0 133.9 9 143.2 143.1 10 37.7 37.7 11 6.98 (8.5) 123.0 122.9 12 6.49 (8.5) 114.3 114.4 13 151.9 152.0 14 131.0 131.0 15 3.28 m 27.1 27.2 16 1.32 d (6.5) 20.3* 20.4 17 1.34 J (7.0) 20.4* 20.4 18 0.90 a 33.2 33.2 19 0.93a 2 1 .6 2 1 .6 2 0 1.60 a 25.2 25.2 OH(13) 4.38 a * interchangeable values From correlations observed in the HMBC experiment, the isopropyl group was placed at position 14 on the aromatic C ring. A V correlation was detected from the H-15 methine to C-14, with a V correlation to C-13 and the isopropyl methyl groups CH 3 I 6 and CH3 I 7 also displayed V correlations to C-14. H-12 exhibited a V correlation to C-13 which carried the hydroxyl group, and V correlations to the aromatic carbons C-9 and C- 14. The H-11 methine showed long range coupling to C-10 at the A/B ring junction and to the aromatic carbons C -8 and C-13. The CH 3-2 O displayed correlations to both - 131 - carbons at the A/B ring junction, that is, a V correlation to C-10 and a V correlation to C-5 and also to C-1 on the A ring. Table 3.5.1b: Long-range connectivities for CN-001 detected in an HMBC experiment Position______Correlated C-atom HMBC (H—»C) VV 1 C-10 C-3, C-5 2 C-3 3 C-1, C-5 5 C-4, C-6 , C-10 C-18, C-19 6 C-5, C-7 C-8 , C-10 7 C-6 , C-8 C-5, C-9, C-14 11 C-8 , C-10, C-13 12 C-13 C-9, C-14 13 OH C-13 C-12, C-14 15 C-14, C-16, C-17 C-13 16/17 C-15 C-14 18 C-4 C-3, C-5, C-19 19 C-4 C-3, C-5, C-18 2 0 C-10 C-1, C-5, C-11 The H-5 methine had V correlations to the neighbouring carbons C-4 and C -6 and also to CIO, and displayed V correlations to the methyls CH 3I 8 and CH 3 I9 , which in turn had V correlations to C-4. Both methyl groups showed V correlations to C-3 and C-5. The H 2-3 methylene exhibited V correlations to C-1 and C-5 and the Hi-2 methylene had a V correlation to C-3. Finally, H 2-I had a V correlation to C-10 and an ^H^^H connectivity was observed in the COSY spectrum between the H 2-I and H 2-2 methylenes. These data were consistent with that reported in the literature for totarol (Nishida et al., 1977; Ying and Kubo, 1991) and compound CN-001 was therefore assigned as totarol (8,11,13-totaratrien-13-ol). - 132- o in 00 r-4 Tf CNO^inT-4'^fnr\]C\o^r~'r40M(NCOvovooo^fnom OH 16/ 17 20 18/ 19 11 12 7a+b Ib 2b2a/6 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm Figure 3.5.1b: *F1 spectrum for CN-001 The stereochemistry of totarol was determined by examination of the 'H and NOESY spectra. The magnitude of the coupling constant for H-5 (J= 12.5 Hz) indicated an axial (a) orientation as described in the literature. A NOE correlation was seen between H-5 and the CH 3-18 methyl indicating that they must be on the same face of the molecule. A further correlation was seen between the CH 3-2 O and CH 3-I9 methyls which were assigned an axial (3 orientation, giving a trans A/B ring junction. - 133 - ppm OCÜ 1.0 - 1 .2 - 1.4- 1 .6 - 1 .8 - 2 .0 - 2 .2 - 2.4 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1.0 0.9 ppm Figure 3.5.1c: NOESY spectrum of CN-001 - 134- 3.5.2 CN-002 (7a-Hydroxytotarol) CN-002 was isolated as a clear oil. The GC-MS data gave a molecular ion at 302 m/z with peaks at 284 m/z and 269 m/z suggesting the loss of water and water plus a methyl group respectively. The proton and carbon spectra were very similar to those of CN- 001. As for CN-001, the carbon spectrum had six signals in the aromatic region, with two doublets in the proton spectrum at 5 7.01 and 5 6.61 each integrating for one proton, which suggested the protons were ortho to each other on the aromatic ring. The two main differences between the carbon spectra of the two compounds were in the resonances of carbons 6 and 7. In CN-001, C-7 resonated at ô 28.7 as a methylene, but in CN-002 it was found downfield at 5 65.6 and was determined from the DEPT-135 experiment as a methine, suggesting the presence of a hydroxyl group attached to C-7. This was also confirmed by the difference in molecular weight between the compounds of 16 Da. The CH 2-6 methylene also resonated downfield in CN-002 at 5 29.1 compared with 5 19.4 for CN-001. This also suggested the presence of a hydroxyl on C- 7 which would probably cause deshielding of C- 6 . The proton spectrum of CN-002 had two singlets at ô 4.98 and ô 4.55 which each integrated for one proton. The signal at 5 4.55 was attributed to a hydroxyl on the aromatic ring, as seen for CN-001. The other broad singlet at Ô 4.98 was determined from the HMQC experiment as the proton attached to C-7. OH Figure 3.5.2a: CN-002 OH -1 3 5 - Figure 3.5.2b: GC trace of CN-001 TIC 141?04r«0*.0 Figure 3.5.2c: GC trace of CN-002 5 00 3000 35 00 4000 45 OC 50 00 55 00 - 136- Unknown: Scan 7247 (40.022 mink 141204nw07.D 100 - 271 175 50 201 286 189 159 115128 215 229 243 0 255 0 20 40 60 80 100 120 140 180 180 200 220 240 260 280 300 Figure 3.5.2d: Mass spectrum of CN-001 Unknown: Averaae of 42.164 to 42.216 min.: 141204nw04.D 269 100^ 227 199 157 50 284 302 I 55 69 83 115128141 241 259 I— 11, [ I I m 11111 n I I'll I lYl'i I I'lYi'i 11 ir [‘lVri^ ‘h *iV|iYr n V i¥TY*'’ lY n ' n T n T q r n r n ...... i | 11111111111111111■[Tn j I iVn*i n I'lmV 111 w m n 111 m t m 111 >111n I n 11 ri 11 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 Figure 3.5.2e: Mass spectrum of CN-002 - 137- TaHe 3.5.2a: (500 MHz) and (125 MHz) NMR spectral data for CN-002 in c d : i3 Postion 'H (yin Hz) '"C ‘^C (Evans et al., 2000b) 1 1.34 m 39.1 39.2 2.20 brd{\2.5) 2 1.57 m 19.5 19.5 l.lA d t 3 1 .2 0 m 41.4 41.5 1.49 m 4 3Z8 32.9 5 1.63 ûfJ (13.0, 2.0) 44.1 44.2 6 1.89 (/r (13.0, 4.0) 29.1 29.1 1.99 brd{U .5) 7 4.98 5 65.6 65.7 8 134.6 134.6 9 142.9 142.9 1 0 38.4 38.4 11 7.01 d (8.5) 123.5 123.5 1 2 6.61 i/( 8 .0 ) 117.1 117.2 13 152.8 153.0 14 133.0 133.1 15 3.56 m 27.6 27.7 16 1.36 d (7.5) 2 0 .8 * 2 0 .8 * 17 1.41 d{1.0) 2 1 .0 * 2 1 .0 * 18 0.97 5 33.1 33.1 19 0.915 21.7 21.7 2 0 I.IO 5 24.5 24.6 OH 13) 4.55s * interchangeable values The data were in close agreement with that found for totarol, but with an extra hydroxyl groip, located at C-7. The data were also in accordance with that reported for 7a- hydroxytotarol by Evans et al. (2000b). CN-002 was therefore determined as 7a- hydroxytotarol (8,11,13-totaratrien-7,13-diol). The stereochemistry of CN-002 was determined from the *H and NOESY spectra. Examination of the ^H spectrum showed a large coupling constant {J= 13.0 Hz) for the metiine proton H-5 coupling to the axial proton at C- 6 . From the NOESY spectrum, a through space correlation could be seen between H-5 and the CH 3-I 8 methyl indicating - 138- they were on the same face of the molecule and this relative stereochemistry was therefore assigned an axial (a) orientation. A further NOE was noted between the CH 3- 19 and CH 3-2 O methyl group which were assigned an axial (p) orientation and confirmed that the A/B ring was again trans, as seen for CN-001. A correlation was also observed between the proton attached to C-7 and the H-15 methine. The C-7 proton resonated as a broad singlet in the spectrum and did not show any discernable coupling; a large coupling to the H 2-6 axial proton would have been observed if H-7 had an axial position therefore, it was assigned an equatorial (P) orientation. This would place the C-7 hydroxyl in an a position, as determined by X-ray crystallography studies on this compound by Pettit et al. (2004), which established the C-7 hydroxyl group as a. ppm - 17 1.0 - 1 .2 - 1.4 - 1.6 - 1 .8 - 2 .0 - Figure3.5.2f: NOESY spectrum of CN-002 2.2 - 2.4 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1.0 0.9 ppm - 139- p p m 1.0 - #' 'It a 1 . 5 - 2 .0 - 2 . 5 - 3 . 0 - Figure 3.5.2g: NOE between 3 . 5 - H-7 and H-15 of 4 . 0 - CN-002 4 . 5 - 5 . 0 - 5 . 5 - 6 .0 - 6 . 5 - 7 . 0 - 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm uir-ii_)vouJi—ilN'q’r.iuiOOvnjOVOOV^T—tVOiHvJit^rnOOI o VO œ o cq ^ oooot^THTHOvoovoininiNïN'q’rot-i.HrHoaNCor'VOrN lTi iTï in fN oocnovovcvooooooooMr^r-voiovovoiovoininininin OH 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm § S' Figure 3.5.2h: *H spectrum of CN-002 - 140- 3.5.3 CN-003 (Sempervirol) Compound CN-003 was isolated as a clear oil. The GC-MS data gave a molecular ion at m/z 286, with a large peak at 271 m/z suggesting the loss of a methyl group and a ftirther peak at 243 m/z indicating the loss of an isopropyl group. The ^^C spectrum solved for 2 0 carbons, six of which resonated in the aromatic region of the spectrum and from the DEPT-135 four were found to be quaternary. The highest resonance was at ô 150.2, characteristic of a phenolic hydroxyl. The data suggested that the CN-003 was a phenolic diterpene and was very similar to that for ferruginol. Two singlets were observed in the spectrum at 5 7.05 and ô 6.40, each integrating for one proton, indicating that the protons were para to each other on the substituted ring. A septet at 3.11 integrating for one proton and two doublets at 5 1.23 and ô 1.21, each integrating for three protons, were again characteristic of an isopropyl group as seen in CN-001 and CN-002. OH Figure 3.5.3a: CN-003 - 141 - Table 3.5.3a: (500 MHz) and (125 MHz) NMR spectral data for CN-003 in CDCI3 Position 'H (Jin Hz) '"C 1 1.36 m 39.1 2.26 b rd (12.0) 2 1.59 m 19.4 1.12 d t (14.0, 3.5) 3 1.24 m 41.7 1.45 ûfJ (13.5, 1.5) 4 33.4 5 1.30 (12.0, 2) 50.7 6 1.66 m 19.1 1.82 m 7 2.80 m 30.0 8 133.8 9 142.9 10 37.5 11 7.05 s 122.4 12 131.7 13 150.2 14 6.40 s 114.9 15 3.11 sep 27.3 16 1.21 d(1.0) 22.7* 17 1.23 J (7.0) 22.8* 18 0.92 5 33.3 19 0.90 5 21.6 20 1.155 25.0 OH(13) 4.40 5 * interchangeable values On first inspection, the *H spectrum appeared to be that of ferruginol, but with the resonances for C-12 and C-13 transposed. However, from ^H-^^^C connectivities observed in the HMBC experiment it could seen that the position of the hydroxyl and isopropyl groups on the aromatic ring were different from ferruginol. The H-15 methine showed a V correlation to C-12 and V correlations to C-11 and C-13 and to the isopropyl methyls CH 3-I 6 and CH 3-I 7 , which in turn had V connectivities to C-15 and V correlations to C-12. This placed the isopropyl group at position 12 on the C ring, not position 13 as seen in ferruginol. - 142 Table 3.5.3b: Long-range connectivities for CN-003 detected in an HMBC experiment Position______Correlated C-atom HMBC (H ^C) V 1 -- 2 -- 3 -- 5 C-4, C-6 , C-10 c-19, C - 2 0 6 C-5 C-10 7 C-6 , C-8 C-5, C-9 11 - C-8 , C-10, C-13, C-15 13 OH C-13 C-12, C-14 14 C-13 C-7, C-9, C-12 15 C-12, C-16, C-17 C-11, C-13 16/17 C-15 C-12 18 C-4 C-3, C-5, C-19 19 C-4 C-3, C-5, C-18 2 0 C-10 C-1, C-5, C-11 The proton of the hydroxyl group displayed a V correlation to C-13 and V correlations to C-12 and C-14 which determined the hydroxyl as being attached to C-13. This revealed that in CN-003, the hydroxyl and isopropyl groups were transposed in comparison with ferruginol. To complete the aromatic ring, the H-14 methine had a V connectivity to C-13 and V connectivities to C-7, C-9 and C-12. The H-11 methine had Vcorrelations to C-8 , CIO, C-13 and C-15. On the B ring, the H]-? methylene showed V correlations to C -6 and C -8 and V correlations to C-5 and C-9. The H 2 -6 methylene exhibited correlations to C-5 (^J) and C-10 (^J) at the A/B ring junction. The H-5 methine displayed V connectivities to C-4, C-6 and C-10, with V connectivities to the methyls CH 3-I 8 and CH 3-I 9 . These methyls in turn had V correlations to C-4 and V correlations to C-3 and C-5. The H 2-I, H2-2 and H 2-3 methylenes of the A ring did not display any long range connectivities in the HMBC experiment, however, in the COSY spectrum correlations - 143- were observed between the H 2-I and H ]-2 methylenes and H 2-I and H 2-3 methylenes. The position of C-3 was further determined by the V correlations displayed by the methyls attached to C-4, and the position of C-1 was confirmed by a V connectivity displayed by the CH 3-2 O methyl, which also had connectivities to C-5 and C-9 and a V connectivity to C-10. Due to the similarity of CN-003 to ferruginol and from comparison with data reported by Mangoni and Caputo (1967), CN-003 was assigned as sempervirol (phenanthren- 2 -ol). t W M TT I O ^ 00 ^ ^ ^ VO 16/ 17 18/ OH 7 . 0 6 . 5 6.0 5 . 5 5.0 4 .5 4.0 3 .5 3 .0 2.5 2.0 1.5 1.0 p p m 1:1 laf 1:1 Figure 3.5.3b: spectrum of CN-003 A large coupling constant {J= 12.0 Hz) was again seen in the spectrum for the H-5 methine, which determined an axial orientation for this proton. The relative stereochemistry was determined from the NOESY spectrum. An interaction was seen between the H-5 methine and CH 3-18 methyl which was assigned an (a) orientation. An NOE was observed between the CH 3-I 9 and CH 3-2 O methyls which were assigned an axial (P) orientation as described in the literature. - 144- 3.5.4 CN-004 (Sugiol) Compound CN-004 was isolated as a clear oil. The GC-MS data gave a molecular ion of 300 m/z, with a large peak at 285 m/z suggesting the loss of a methyl and a further peak at 243 m/z indicative of the loss of an isopropyl group. Examination of the ^^C and DEPT-135 spectra revealed six resonances between 8 110 and 8 157, four of which were quaternary carbons, indicating a ^e^ra-substituted aromatic ring. A further signal at 8 198.4 was characteristic of a ketone. This was confirmed by a singlet seen at 8 7.89 in the ’H spectrum suggesting that it was in close proximity to a carbonyl group. The characteristic signals for an isopropyl group were also detected in the spectrum, with a septet at 8 3.11 which integrated for one proton and doublets at 8 1.24 and 1.26 which together integrated for six protons. The ^^C spectrum showed 20 carbons and the data suggested that the compound was an aromatic abietane diterpene containing a ketone. The singlet at 8 7.89 and another at 8 6 .6 6 indicated that the protons did not couple and were para to each other, therefore the compound was likely to be of the ferruginol type rather than a totarane. OH Figure 3.5.4a: CN-004 - 145- Table 3.5.4a: *H (500 MHz) and (125 MHz) NMR spectral data for CN-004 in CDCI3 Position 'H(/in Hz) '"C ^^C (Wenkertet al. 1976) 1 1.52 j 37.9 37.3 2.2 brd{\2.Qi) 2 1.66 (/r (14.0, 3.5) 18.9 18.4 1.76 (13.5, 3.0) 3 1.25 m 41.4 40.9 1.50 s 4 33.2 32.7 5 1.83 c/û? (14.0, 4.0) 49.5 49.1 6 2.57 (18.0, 13.5) 36.1 35.4 2.66 (18.0, 4.0) 7 198.4 199.2 8 124.8 122.4 9 156.4 156.3 10 37.9 37.4 11 6.66 s 110.0 109.0 12 157.2 160.3 13 132.0 133.0 14 7.89 5 126.6 125.6 15 3.11 sep 2&8 26.1 16 1.24 d (7.0) 22.3* 21.6* 17 1.26 d (7.0) 22.4* 21.8* 18 0.915 32.6 31.9 19 0.97 5 21.4 20.7 20 1.21 5 23.2 22.5 OH(12) 5.15 5 * interchangeable values Comparison of the ^^C spectrum with that for ferruginol revealed that the main differences between the two compounds were for C-6 and C-7. In CN-004, C-6 had a 5 value of 35.4, whereas in ferruginol C-6 resonated upheld at ô 19.2. The biggest difference in resonance was for C-7, which in CN-004 was found at 5 199.2, but was seen at 5 29.6 in ferruginol. This suggested that C-7 was a ketonic carbon in CN-004 and that the resonance of C-6 was deshielded by the sp^ carbonyl group, resulting in a higher ô value than seen in ferruginol. From the HMBC experiment, the isopropyl - 146- methine H-15 showed a V correlation to C-13, with the isopropyl methyl groups (C-16, C-17) showing V correlations to C-13. The methine H-15 also had V correlations to C- 12 and C-14, which placed the isopropyl group at position 13 on the C ring. The H-11 methine displayed a V correlation to C-12 and V connectivities to C -8 and C-13. The H-14 methine had V connectivities to C-7, C-9, C-12 and C-15, completing the assignment for the aromatic ring. Table 3.5.413.5.4b: Long-range ^H—>^^C cormectivities for CN-004 detected in an HMBC experiment Position Correlated C-atom HMBC (H—>C) 1 - - 2 -- 3 -- 5 - - 6 C-5, C-7 - 11 C-12 C-8 , C-13 12 OH - C-13 14 - C-7, C-9, C-12, C-15 15 C-13, C-16, C-17 C-12, C-14 16/17 C-15 C-13 18 C-4 C-3, C-5, C-19 19 C-4 C-3, C-5, C-18 2 0 C-10 C-1, C-5, C-11 The ketone was placed at C-7 on the B ring, due to a V correlation displayed by the H 2 -6 methylene and a V correlation by the H-14 methine. The HMBC experiment did not reveal any long-range ^H to ^^C connectivities for the methylenes H 2-I, H2-2 , or H2-3 on the A ring or for the H-5 methine at the A/B ring junction. This was probably due to the small amount of sample (<1.5 mg) available for the experiment. However, long-range coupling data for the methyl groups (C-18, C-19, C-20) and for the H 2 -6 methylene, plus information from the COSY and NOESY spectra facilitated assignment of the A ring carbons. The two methyls (C-18, C-19) exhibited a V connectivity to the quaternary C- - 147- 4 and Vconnectivities to C-3 and C-5. The H2 - 6 methylene also showed a Vcorrelation to C-5. The CH 3-2 O methyl showed a V connectivity to C-10 and Vconnectivities to C- 1, C-5 and C-11. From the COSY experiment, a signal was detected between the H-5 methine and the H2 - 6 methylene, and between the H 2-I and H 2-2 methylenes. In the NOESY experiment, a through space interaction was seen between the H-11 methine and H 2-I methylene which completed the assignment of the A ring. These data were compared with those reported in the literature (Chang et al., 1990; Wenkert et al., 1976), and compound CN-004 was determined as sugiol ( 12-hydroxy-8,11,13-abietatrien-7- one). Ln«—tr^'^cocoocr»cNro'X>ovx»«-HroroLnro(Nv-D rO rO rO CN (N CN CN (N 18/ 19 11 14 OH J U J iLxJWUuuLl ' I ' ' ' ' I ' ' ' ' I ' ' ' ' I ' ' ' ■ I ' ' ' ' I ' ' ' ’ I ' ' ' ' I ' ' ' ' I ' ' ' ' I ' ' ' ' I ' ' ' ' I ' ' ' ' I ' ' ' ' I ' ■ ■ ' I ' ■ - ' 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm 00tN CMT- ftO Oh- CO Figure 3.5.4b: ’H spectrum of CN-004 - 148- From the spectrum, the H-5 methine had a large coupling {J= 14.0 Hz) to the axial proton at C- 6 . In the NOESY experiment, a through space interaction was seen between the H-5 methine and the CH 3-I 8 methyl which again showed they were on the same face of the molecule and they were assigned an a orientation as reported in the literature. The CH 3-2 O methyl had an NOE to the CH 3-I 9 methyl and they were again assigned a (P) axial orientation. 149- 3.6 PINUS NIGRA 3.6.1 Preliminary Screening of Six Pinus species The immature cones of Pinus muricata, P. nigra, P. nigra var. larico, P. pungens, P. radiata and P. sylvestris were investigated for anti-staphylococcal activity. Cold solvent extraction starting with hexane, followed by EtOAc then MeOH yielded crude extracts which were tested in the MIC assay. The MIC results revealed that as with the other conifer species screened, the hexane extracts were the most active. The general trend was that the EtOAc extracts were the next most active with the MeOH extracts being the least active. Although, for P. nigra and P. nigra var. larico the hexane and EtOAc extracts were equally active and with P. pungens and P. sylvestris, the MeOH extracts were as active or more active than the EtOAc extracts. The results suggested that further investigation of cones from any of thePinus species tested would have been worthwhile, however, the good supply P.of nigra cones and the anti-HIV activity reported for a P. nigra cone isolate (Eberhardt and Young, 1996) determined the selection of these cones for further fractionation. Figure 3.6.1: TEC of Pinus species hexane extracts (sprayed with vanillin/H2S0 4) M = P. muricata ? = P. pungens N L = P. nigra var. larico N = P. nigra AA = commercial abietic acid R = P. radiata I i ■ I S = P. sylvestris M P N L N A A R S - 1 5 0 - Table 3.6.1: MICs (p,g/ml) of cone extracts jfrom Pinus species against S. aureus Pinus species Extract ATCC 29523 XU212 (TetK) Pinus muricata Hexane 64 64 EtOAc 256 256 MeOH 512 512 Pinus nigra Hexane 128 64 46 44 EtOAc 128 64 44 44 MeOH 256 256 Pinus nigra var. larico Hexane 128 64 EtOAc 128 128 MeOH 128 256 Pinus pungens Hexane 64 64 EtOAc 512 512 MeOH 256 256 Pinus radiata Hexane 256 256 EtOAc 512 512 MeOH 512 512 Pinus sylvestris Hexane 128 128 EtOAc 256 256 MeOH 256 128 Tetracycline 0.25 128 - 151 - 3.6.2 Extraction of PN-001 fromPinus nigra Cold solvent extraction on 140 g immatureP. nigra cones yielded 4.2 g crude hexane extract. This extract was subjected to VLC and eluted with a gradient of 10% increments, starting with 100% hexane to 100% EtOAc. Finally, the colunrn was washed with EtOAc/MeOH 50:50. Biotage^^ flash chromatography was performed on 475 mg of the most active VLC fraction (4). A silica column was used and elution commenced with 100% hexane, employing a gradient of 20% increments to 100% EtOAc. A sample of each fraction was run on TLC and fractions with a similar profile were pooled (9-12, 19-24, 27-47, 48-60, 62-79). The active fraction (27-47) was then further purified by PTLC (toluene/EtOAc/AcOH 90:80:2, two developments), which yielded a colourless crystalline solid (16 mg). VLC4 1-8 9-12 13-17 18 19-24 25 26 27-47 48-60 61 62-79 Figure 3.6.2: TLC of pooled flash chromatography fractions (sprayed with vanillin/H2S04) - 152- 3.6.3 PN-001 (Isopimaric acid) Compound PN-001 was isolated as a white crystalline solid. The GC-MS data gave a molecular ion at 302 m/z. The experiment solved for 20 carbons, suggesting that the compound was a diterpene. A signal at Ô 185.0 was characteristic of a carboxylic acid, with two peaks in the IR spectrum at 2925 and 1689 cm'^ also indicative of a carboxylic acid functional group. Four other signals were seen in the aromatic/olefinic region of the carbon spectrum, one of which at 5 109.3 was determined from the DEPT-135 experiment as a methylene, indicating that the compound contained an exomethylene group. The data suggested a molecular formula of C 20H30O2 and that PN-001 might be a resin acid. 17 16 15 Figure 3.6.3a: PN-001 The data were similar to those of Wenkert et al. (1972) suggestive of an isopimarane diterpene. Signals in the experiment at 5 5.78 (H-15) and 5 4.85 and 4.91 (H 2-I 6 ) for a vinyl group supported this. 153 Table 3.6.3a: ‘H (500 MHz) and (125 MHz) NMR spectral data for PN-001 in CDCI3 Position 'H (yin Hz) (Wenkert e/fl/. 1972) 1 0.99 m 38.8 3&2 1.83 m 2 1.54 m 17.9 17.9 3 1.65 m 37.0 37.2 4 46.3 46.4 5 2.2\dd(\2.5, 2.0) 45.0 45.4 6 1.69 m 25.2 25.7 1.93 J (3.0) 7 5.31 r(4.0, 1.5) 121.0 121.5 8 135.7 136.0 9 1.75 m 52.0 52.4 10 35.0 35.5 11 1.34 m 20.0 20.5 12 1.41 m 36.1 36.0 1.46 m 13 36.8 37.5 14 1.90 m 46.1 46.5 1.99 m 15 5.78 c/û? (17.5, 11.0) 150.3 150.7 16 4.85 (11.0, 1.5) 109.3 109.7 4.9 (17.5, 1.5) 17 0.84 5 21.5 21.9 18 185.0 183.9 19 1.26 5 17.1 17.5 20 &895 15.3 15.7 In the ^H experiment, the splitting pattern seen for the protons attached to C-15 and C-16 revealed that these carbons formed part of a vinyl group. The H-15 methine resonated as a double doublet as its signal was split by the two neighbouring protons which were each in a different chemical environment. The J coupling constants for the H-16 methylene revealed a cis coupling (/= 11.0 Hz) between H-16a and H-15, both protons being on the same side of the double bond, and a geminal coupling to H-16b {J = 1.5 Hz). Proton H-16b was trans to H-15, revealed by its large J value of 17.5 Hz. The HMBC experiment showed a V correlation from the H 2-I 6 methylene to C-15 and a V - 154 correlation to C-13. The H-15 methine had a J connectivity to C-13 and V connectivities to C-12 and C-14 and to the CH 3-I 7 methyl. This data indicated that the exomethylene group was attached to the quaternary C-13. The CH 3-I 7 methyl also displayed a V correlation to C-13 and V correlations to C-14 and 15. The position of C- 12 was determined by V connectivities from the H 2-I 2 methylene to C-11 and C-13 and V connectivities to C-9 and C-14. The H 2-II methylene showed a V correlation to C-9 and V correlation to C-13. The position of C-14 was confirmed by V correlations from the H 2-I4 methylene to C -8 and C-13 and V correlations to C-7, C-9, C-12 and C-17. This completed the assignments for the C ring. Table 3.6.3b: Long-range connectivities for PN-001 detected in an HMBC experiment Position______Correlated C-atom HMBC (H—^C) VV 1 C-10 C-5, C-20 2 C-3 C-4 3 C-2, C-4 C-1,C-19 5 C-6 C-3, C-18 6 C-7 C-4, C-8 , C-10 7 C-6 C-5, C-9 9 C -8 C-14 11 C-9 C-13 12 (M l, C-13 C-9, C-14 14 C-8,C-13 C-7, C-9, C-12, C-17 15 C-13 C-12, C-14, C-17 16 C-15 C-13 17 C-13 C-14, C-15 19 C-4 C-3, C-5 2 0 C-10 C-1, C-5, C-9 An additional olefrnic proton could be assigned as H-7 on the B ring, on the basis of a V correlation to C-6 and V correlations to C-5 and C-9. The H 2 -6 methylene showed a V connectivity to C-7 and V connectivities to the olefrnic C -8 and to C-4 and C-10. The - 155- COSY spectrum revealed a coupling between H 2 -6 to a methine (H-5) which in the HMBC spectrum showed a V correlation to C -6 and V correlations to C-3 on the A ring and to C-18 of the carboxylic acid. A V correlation from the CH 3-2 O methyl to C-5 also determined the position of C-5 at the A/B ring junction. Additional couplings could be seen to the quaternary C-4 with V correlations from the CH 3-I 9 methyl and the H 2-3 methylene and a V correlation from the H-2 methylene. The CH 3-I 9 methyl protons also displayed V connectivities to C-3 and C-5, and the H 2-3 methylene showed connectivities to C-19 (^J) and to C -2 (V ) which in turn had a V connectivity to C-3. The position of C-1 on the A ring was established by correlations from the H 2-I methylene to C-5 and C-10 at the A/B ring junction and to C-20. The CH 3-2 O methyl also showed correlations to both carbons at the A/B ring junction and also to C-1 and C- 9. The similarity of these data to those reported by Wenkert et al. (1972) led to the assignment of PN-001 as isopimaric acid (7,15-isopimaradien-18-oic acid). The relative stereochemistry of isopimaric acid was determined from coupling data and from the NOES Y experiment. The large coupling constant {J = 12.5 Hz) for H-5 established an axial (a) orientation for H-5. From the NOES Y spectrum, a through space correlation was seen between the methyl groups CH 3-2 O and CH 3-I 9 which established that they were on the same face of the molecule and were given an axial (P) orientation as described in the literature. The CH 3-2 O also had a NOE correlation to one of the H 2-II protons, which in turn had a NOE 1—>3 interaction to the CH 3-I 7 methyl group. The H 2-II methylene resonated as a multiplet in the *H experiment and therefore did not provide any coupling data. However, since the CH 3-2 O methyl group had an axial orientation, the NOE correlation must be to the axial proton of H 2-II. It was - 156- therefore considered likely that the CH3-I7 methyl would also been in an axial orientation in order to have an NOE interaction to the axial proton on C-11. This determined the relative stereochemistry of the C-15/C-16 exomethylene group as equatorial and (a). ( N J O r ^ L D i - i r O O r ^ ’cj'fNJCTsOOOO^'X) v£>^^OOmLOOr^jr^»HCT\rHrHO'^r-H'^COOmOOLn(N'>DO^'J3(NOtnC^ t-»a>r'UirHOO(NCNa>oo'^LDroro t^»^rorvjo 16b 16a 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm § S Figure 3.6.3b: ’H spectrum for PN-001 - 157- ppm 0 .6 - 0 .8 - 1.0 - 1 .2 - 1.4- 1.6 - 1.8 - 2 .0 - 2 .2 - 2 .4 - 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 ppm Figure 3.6.3c: NOESY spectrum of PN-001 - 158- 4.0 MIC AND MODULATION ASSAYS - RESULTS AND DISCUSSION 4.1 RESULTS OF ANTI-STAPHYLOCOCCAL MIC ASSAYS All isolated compounds were tested in the minimum inhibitory concentration (MIC) assays as described in Section 2.3.1. Six S. aureus strains were used: standard strain ATCC 25923, three effluxing strains XU212 (TetK), SA1199B (NorA), RN4220 (MsrA) and two epidemic MRS A strains EMRSA-15 and EMRSA-16. All strains with the exception of ATCC 25923 are clinically relevant isolates. For ease of reference the table (from Section 2.3.3) of MICs for standard antibiotics against these strains is repeated below. Table 4.1a: MICs (pg/ml and pM) for standard antibiotics tested against S. aureus strains used in MIC and modulation assays S, aureus tetracycline norfloxacin erythromycin oxacillin strain R >16pg/mi R >16pg/ml R >8pg/ml R >4pg/ml ATCC 25923 0.25 1 0.25 0.125 XU212 128 16 4,096 256 SA1199B 0.25 32 0.25 0.25 RN4220 0.25 2 128 0.25 EMRSA-15 0.125 0.5 2,048 32 EMRSA-16 0.125 128 4,096 512 S, aureus tetracycline norfloxacin erythromycin oxacillin strain R>33.3pM R>50.1pM R>10.9jaM R>9.1pM ATCC 25923 0.5 3.1 0.3 0.3 XU212 266.2 50.1 5,581.1 580.0 SA1199B 0.5 1 0 0 .2 0.3 0 .6 RN4220 0.5 6.3 174.4 0 .6 EMRSA-15 0.3 1 .6 2,790.6 72.5 EMRSA-16 0.3 400.9 5,581.1 1,159.9 R = MIC value which indicates antibiotic resistance It can be seen from the above table that these strains are highly resistant to some standard antibiotics. There is very considerable resistance to erythromycin shown by both EMRSA strains and also XU212. In fact, XU2I2 is highly resistant to - 159- tetracycline, erythromycin and oxacillin and exhibits borderline resistance to norfloxacin. The epidemic strain EMRSA-16 also shows resistance to oxacillin, erythromycin and norfloxacin, but is susceptible to tetracycline. Abietic acid is a diterpene resin acid found in Pinus species and a sample (70% purity) was obtained from Aldrich. Abietic acid was tested in the MIC and modulation assays as a comparison for isopimaric acid. Table 4.1b: MICs (pg/ml) of isolated compounds against a standard ATCC strain and effluxing strains of S. aureus Compound ATCC XU212 SA1199B RN4220 25923 (TetK) (NorA) (MsrA) Ferruginol 8 8 4 8 Formosanoxide >512 >512 n/t n/t 4p-Hydroxygermacra- 128 128 256 128 l(10)-5-diene Isopimaric acid 64 32 32 64 Oplopanonyl acetate >128 >128 >128 >128 Pisiferol 16 8-16 8-16 8 5-Epipisiferol 8 16 8 16 Sempervirol 2 2 2 2 Totarol 2 2 2 2 7a-Hydroxytotarol 8 16 8-16 16 Torulosal >128 >128 >128 >128 ^ra«5 -Communic acid 16 8-16 8 8 Sugiol n/t > 1 0 > 1 0 > 1 0 Abietic acid 64 64 64 64 n/t = not tested - 160- It could be seen that the phenolic diterpenes had the greatest activity against the effluxing strains, with totarol and sempervirol being the most active compounds followed by ferruginol. 7a-Hydroxytotarol only differs from totarol by an extra hydroxyl group, but was 4 to 16 times less active, perhaps due to the increase in hydrophilicity caused by the additional hydroxyl group, which may affect the ability of the compound to cross the lipophilic bacterial cell membrane. Pisiferol and 5- epipisiferol had the same activity, therefore the difference in stereochemistry at the A/B ring junction did not appear to have an effect on antibacterial activity. Table 4.1c: MICs (pg/ml) of isolated compounds against EMRSA strains Compound EMRSA-15 EMRSA-16 Ferruginol 16 4 Formosanoxide n/t n/t 4p-Hydroxygermacra-1 ( 10)-5-diene 256 128 Isopimaric acid 32 64 Oplopanonyl acetate >128 >128 Pisiferol 8 8 5-Epipisiferol 16 8 Sempervirol 2 2 Totarol 2 1 7a-Hydroxytotarol 16 8 Torulosal >128 128 ?ra«5 -Communic acid 16 8 Sugiol > 1 0 n/t Abietic acid 64 64 n/t = not tested - 161 - Totarol was also the most active compound against the EMRSA strains with MICs of 2 and 1 pg/ml against strains 15 and 16 respectively. Sempervirol also had very good activity of 2 pg/ml against both strains. Ferruginol, pisiferol, 5-epipisiferol and 7a-hydroxytotarol showed good activity compared to the standard antibiotics. rra«5 -communic acid had greater activity than isopimaric acid and abietic acid against both the effluxing and MRS A strains, with MICs of 8-16 pg/ml compared with 32-64 pg/ml. Sugiol was extracted from C. nootkatensis in ver>^ small quantities of less than 0.5 mg per HPLC run. Due to a shortage of material, it was decided to test sugiol at 10 pg/ml in the modulation assay. This served a dual function, if the compound was antibacterial at this concentration, then the growth control would be inhibited, indicating good antibacterial activity of less than 10 pg/ml. If, however, the growth control was unaffected, indicating no antibacterial activity at 1 0 pg/ml, it was possible that the compound might be a potentiator of antibiotic activity and this would be revealed in the modulation assay. Sugiol did not inhibit the growth control at this concentration, therefore, if it does possess anti-staphylococcal activity it must be at a greater concentration than 10 pg/ml. The compound also did not exhibit any modulating activity. It is interesting to note that nearly all the compounds had a higher MIC against EMRSA-15 than for EMRSA-16. This was surprising since the MIC for oxacillin against EMRSA-15 is only 32 pg/ml compared with 512 pg/ml against strain 16. EMRSA-16 is resistant to norfloxacin where strain 15 is not and the MIC for erythromycin is also two-fold greater for EMRSA-16 than against strain 15. 162 Torulosal was inactive at 128 |ag/ml against all strains except for EMRSA-16, where the MIC was 128 pg/ml. It is possible that acquisition of resistance to several antibiotics has resulted in a loss in fitness, making EMRSA-16 more susceptible to some compounds or types of compounds than EMRSA-15. In culturing the S.aureus strains many times on agar slopes, it was noticeable that strain ATCC 25923 in particular grows much better than EMRSA-16. Another possibility is that EMRSA- 15 may possess an as yet uncharacterised mechanism of resistance, or some efflux of compounds may occur, accounting for the higher MICs seen for this strain compared with EMRSA-16. There was no significant variation in the activity of a compound between the six strains. For example, there were no instances where a compound had a high activity against the standard ATCC 25923 strain, but considerably reduced or even no activity against the resistant strains. None of the compounds showed activity against the effluxing strains but not against the EMRSA strains or vice versa. With the exception of ferruginol, which had a four-fold higher activity against EMRSA-16 than strain 15, none of the compounds had more than a two-fold difference in activity between strains. This was a surprising result which may reflect that the active compounds were all diterpenes, some with a very similar structure and with the same functional groups and possibly the same mode of action. However, trans-œmmumc acid, which had good activity compared with the standard antibiotics, is a labdane diterpene and does not possess a phenolic hydroxyl group as found in the active abietanes. Isopimaric acid and abietic acid, although not as active, still displayed the same trend, that if they were active against one strain, they were active against all strains at around the same MIC value. This suggests there may be a common mode -1 6 3 - of action for each compound against all strains, for example membrane perturbation is often suggested as the antibacterial action for phenolic diterpenes. However, research by Clarkson et al. (2003a) demonstrated that ferruginol did not affect erythrocyte morphology even at high concentration (100 pg/ml). Totarol has been shown to act on the bacterial respiratory electron transport chain, inhibiting NADH- cytochrome c reductase (Haraguchi et al., 1996b). However, using isolated rat liver mitochondria, Evans et al. (2000a) determined that totarol did not inhibit respiration at its MIC of 2 pg/ml (7 pM). Concentrations of 50 pM and above were required to inhibit respiration, and it was concluded that interference with respiration is not likely to be the mode of action for the antibacterial activity of totarol and related diterpenes. The results do show, as stated by Evans et al. (2000a), that at its MIC concentration, totarol is not toxic to mammalian mitochondria, which is an important factor for any compound with potential for therapeutic use. A compound which inhibits bacterial respiration but is inactive against the mammalian respiratory chain would be ideal, to avoid potential toxicity. It is possible that some or all of these active compounds have a mode of action that the clinical isolates have not previously encountered. This is one of the factors which make plant secondary metabolites an attractive prospect for developing new antibacterials and is a major driving factor behind this research. 164- 4.2 RESULTS OF MODULATION ASSAYS Modulation assays were carried out on isolated compounds to determine whether they could potentiate antibiotic activity against the effluxing or EMRSA strains as described in Section 2.3.2. Table 4.2a: MICs (pg/ml) of tetracycline, norfloxacin and erythromycin in the presence and absence of isolated compounds ( @ 1 0 pg/ml unless otherwise stated) against effluxing strains of S. aureus XU212 SA1199B RN4220 Compound Tetracycli Norfloxacin Erythromycin ne Reserpine (20 pg/ml) 128 (32) 32(4) 128 (128) Ferruginol (2 pg/ml) 128 (32) 32 (16) 128 (32) Formosanoxide 128 (128) 32 (32) 128 (128) 4p-Hydroxygermacra-1 ( 10)-5- 128 (128) 32 (32) 128 (128) diene Isopimaric acid 128 (128) 32 (32) 128 (128) Oplopanonyl acetate 128 (128) 32 (32) 128 (128) Pisiferol 128 (128) 32 (32) 128 (128) 5-Epipisiferol (8 pg/ml) 128 (32) inhibited control* inhibited control* Sempervirol 128 (128) 32 (32) 128 (128) Totarol (1 pg/ml) 128 (32) 32(8) 128 (16) 7a-Hydroxytotarol (2 pg/ml) 128 (128) 32 (32) 128 (128) Torulosal 128 (128) 32 (32) 128 (128) ^ra«5 -Communic acid 128 (128) 32 (32) 128 (128) Sugiol 128 (128) 32 (32) 128 (128) Abietic acid (20 pg/ml) 128 (128) 32 (32) 128 (128) Figures in bold denote MICs in the presence of the isolated compound * 8 pg/ml 5-epipisiferol inhibited growth; 4 pg/ml had no modulating activity - 165- Only ferruginol and totarol potentiated antibiotic activity against the three effluxing strains. Against XU212, both compounds resulted in a four-fold reduction in the MIC of tetracycline, which was the same reduction seen with reserpine. Totarol was slightly more active than ferruginol against SA1199B and RN4220, resulting in a four-fold and eight-fold increase in antibiotic activity respectively, compared with a two-fold and four-fold increase for ferruginol. Both compounds potentiated the activity of erythromycin against RN4220 which has the MsrA efflux pump, where reserpine is inactive. It is possible that like reserpine, ferruginol and totarol may be efflux pump inhibitors, but this result suggests that they must have a different mode of action from reserpine. One possibility is that reserpine, which is a much larger molecule than totarol and ferruginol, may act by blocking the efflux pump, whilst the diterpenes could exert their effect in other ways such as affecting translation of the (pump) protein or assembly of the pump in the membrane in the correct configuration. No inhibitors of the MsrA pump have been reported and unlike reserpine, totarol and ferruginol have antibacterial activity and it is possible that they may act to inhibit their own efflux from the bacterial cell. Ferruginol and totarol displayed excellent modulatory activity in potentiating the activity of oxacillin against EMRSA-15 (Table 4.2b). Ferruginol reduced the MIC of oxacillin from 32 pg/ml to 0.40 pg/ml, an eighty-fold reduction, which was almost as great as the 1 0 0 -fold reduction seen for epicatechin gallate which reduced the oxacillin MIC from 32 pg/ml to 0.32 pg/ml. Totarol also reduced the MIC of oxacillin from 32 to less than 1 pg/ml and both compounds restored oxacillin sensitivity in this resistant clinical isolate. - 166 Table 4.2b: MICs (lag/ml) of oxacillin in the presence of isolated compounds (@ 10 pg/ml unless otherwise stated) against EMRSA strains EMRSA-15 Compound EMRSA-16 oxacillin oxacillin Epicatechin gallate 32 (0.32) 512(0.16) Ferruginol (2 pg/ml) 32 (0.40) inhibited control# Formosanoxide 32 (32) n/t 4P-Hydroxygermacra- 1(10)-5-diene 32 (32) 256 (256) Isopimaric acid 32 (32) 512(512) Oplopanonyl acetate 32 (64) 512(512) Pisiferol 32 (32) 512(512) 5-Epipisiferol (8 pg/ml) inhibited control* inhibited control* Sempervirol 32 (32) 512(512) Totarol (1 pg/ml) 32 (<1) inhibited control § 7a-Hydroxytotarol (2 pg/ml) 32 (32) 512(512) Torulosal 32 (64) 512 (256) ^ra«5 -Communic acid 32 (32) 512(512) Sugiol 16(16) n/t Abietic acid 32 (32) 512(512) n/t = not tested; #ferruginol @ 1 pg/ml was inactive; * 5-epipisiferol @ 4 pg/ml was inactive; § totarol @ 0.5 pg/ml was inactive Difficulties were encountered with some of the modulation assays, in that a concentration of half the MIC of an active compound inhibited growth right across the plate including the growth control, but at one quarter of the MIC, the compound was inactive and there was no potentiation of antibiotic activity. This was seen with 5-epipisiferol against all strains except XU212 where a concentration of half the MIC caused a four-fold increase in tetracycline activity. This problem was also 167- encountered with ferruginol and totarol when testing for their ability to potentiate oxacillin activity against EMRSA-16. Very good modulating activity was seen for both compounds against EMRSA-15, however, against EMRSA-16, ferruginol at a concentration of half the MIC inhibited the control and at one quarter MIC, was inactive. Of course, compounds which are active at low MICs are highly desirable, but this can lead to difficulties, for example, in the case of totarol, the difference in concentration between an MIC of 1 pg/ml and a half MIC of 0.5 pg/ml is not very great, and any variation in the bacterial titre will have an effect, possibly leading to inhibition of the control. Use of a spectrophotometer to more accurately determine the bacterial titre, rather than reliance on the McFarland Standard might help in situations where compounds have very low MIC values. ‘Pin head’ growth was seen in several wells after using MTT to develop the EMRSA plates which made determination of the MIC difficult or even impossible in some cases, and this was a particular problem with EMRSA-16. It is possible that neither ferruginol nor totarol potentiate oxacillin activity against EMRSA-16. The results suggest that there is a narrow window of modulatory activity for these compounds. This may be partly due to the fact that compounds which were active in the modulation assay also had antibacterial activity and when used at half MIC, there might still be some antibacterial activity. This was in comparison with reserpine which has no anti-staphylococcal activity even at 512 pg/ml and its potentiation of tetracycline and norfloxacin activity would not be due to additive effects. Torulosal had no modulating activity, except against EMRSA-16 where it resulted in a two-fold reduction in the MIC of oxacillin. EMRSA-16 was the only strain against which torulosal had antibacterial activity (at 128 pg/ml) and it is possible that this may account in part for its potentiation of oxacillin activity. However, against 168- EMRSA-15, there was a two-fold decrease in oxacillin activity, but since there is an intrinsic two-fold variability in the results for MIC assays, both these results should be treated with caution. Totarol, at a concentration of 1 ng/ml also reduced the MIC of oxacillin from 256 to 32 |ig/ml against strain XU212. This was an eight-fold increase in antibiotic activity, compared with the four-fold increase in tetracycline activity which was seen when assayed with totarol. The modulating activity of totarol was much greater in potentiating oxacillin activity against EMRSA-15 than was seen against the effluxing strain XU212. It is possible that there may be some efflux of oxacillin occurring in XU212. Sempervirol was inactive in the modulation assays, which was surprising, especially as its activity was almost as good as totarol’s in the MIC assays. Ferruginol, which was less active than sempervirol in the MIC assays, showed good modulating activity. It is possible that the isopropyl group being positioned at carbon 12 in sempervirol, compared with position 13 for ferruginol and 14 for totarol, has an effect on modulating activity. Structure activity relationships will be discussed in Section 4.5. Isopimaric acid did not show any resistance modifying activity. Other isopimaranes have shown moderate potentiation of the activity of tetracycline and erythromycin against XU212 and RN4220 respectively, resulting in a two-fold reduction in MIC for these antibiotics (Gibbons et al., 2003). - 169- 4.3 ASSAYS FOR POTENTIATION OF COMPOUND ACTIVITY 4.3.1 Results Oplopanonyl acetate and torulosal isolated from C lawsoniana, isopimaric acid isolated from Pinus nigra and abietic acid from Aldrich were tested in combination with reserpine against strains XU212 and SA1199B. This was done to see if the presence of an efflux pump inhibitor increased the activity of isopimaric and abietic acid, and in the case of oplopanonyl acetate and torulosal to see whether these compounds had antibacterial activity when the efflux pump was inhibited, as seen with several compounds tested by Tegos et al. (2002). The rationale shown by these authors was that testing an inactive compound in the presence of an efflux pump inhibitor might reveal antibacterial activity for that compound and they suggested that a compound should not be considered inactive until it has been tested with an inhibitor. Table 4.3.1a: MICs (pg/ml) for compounds in the presence and absence of reserpine against effluxing strains of S. aureus Compound XU212 SA1199B (TetK) (NorA) Tetracycline 128 (32) Norfloxacin 32(4) Oplopanonyl acetate 128 (128) 128 (128) Torulosal 128 (128) 128 (128) Isopimaric acid 32 (128) 64 (256) Abietic acid 64 (256) 64 (128-256) Figures in bold denote MICs in the presence of reserpine - 170 Oplopanonyl acetate and torulosal were inactive even in the presence of an efflux pump inhibitor, showing that they do not have anti-staphylococcal activity against these strains. The results for isopimaric acid and abietic acid were surprising; it was expected that if they were substrates for the TetK or NorA efflux pumps, then inhibition of the pump would increase their activity shown by a decrease in MIC. Mtematively, if isopimaric acid and abietic acid were not substrates for these efflux pumps, then inhibition of the pumps would not affect the activity of the compounds and their MIC values would be the same in the presence or absence of reserpine. However, the presence of reserpine resulted in an increase in MIC, indicating a decrease in the activity of both compounds. A four-fold decrease in activity was seen against strain XU212 and a two-four-fold decrease against SA1199B. This unusual result has been seen by other researchers, Tegos et al. (2002) saw a four-fold decrease in the activity of gossypol against a norA strain. Isopimaric acid was tested against EMRSA-15 and abietic acid against both EMRSA strains in combination with epicatechin gallate to see if this compound potentiated Iheir activity. However, the MICs were the same in both the presence and absence of epicatechin gallate. Therefore, unlike reserpine, there was no sign of a potential modulator having a negative effect. 4.3.2 Abietic Acid and Reserpine From the results seen in the MIC assay, it is possible that isopimaric acid and abietic acid may form a complex with reserpine and NMR spectroscopy was carried out to bok for a potential interaction between reserpine and the resin acids. Abietic acid and reserpine (5 mg) were incubated separately, and in combination, in deuterated DMSO and left overnight at 37°C. This was done to replicate, as far as possible, the - 171 - conditions of the MIC assay. Neither abietic acid nor reserpine is soluble in water, so it was not possible to use deuterated H 2O. However, in the MIC assay the compounds were dissolved in DMSO before addition to MHB and DMSO is also more likely to mimic water than chloroform which was normally used to resuspend the compounds. After overnight incubation, NMR was carried out and comparison made of the proton spectra of abietic acid and reserpine individually and as a mixture. Changes in chemical shift of proton resonances of the individual compounds and the mixture were noted (Table 4.3.2a) Table 4.3.2a: Reserpine and abietic acid resonances (recorded at 400 MHz) which showed a change in chemical shift after mixing and overnight incubation Reserpine Reserpine + AA Change in shift (ppm) Hz H 1.949 1.958 0.009 3.6 4a 2.847 2.856 0.009 3.6 6 2.876 2.883 0.007 2.8 5 3.011 3.024 0.013 5.2 7 4.342 4.361 0.019 7.6 13b 10.539 10.551 0 .0 1 2 4.8 13 Abietic Acid 0.746 0.742 -0.004 - 1 .6 2 0 5.712 5.708 -0.004 - 1 .6 14 Reserpine had the greater number of protons exhibiting a change in chemical shift and the changes in Hz were greater than those seen for abietic acid. There may have been further changes in chemical environment of other protons on abietic acid, but there was heavy overlap in the alkyl region of the spectrum for the mixture which made it difficult to detect chemical shift changes in that region. It is possible that seme of the changes in chemical shift may be due to a conformational change in reserpine when interacting with abietic acid. - 172- Reserpine MeO +4.8 H OMe O* OMe Abietic acid OMe OMe Figure 4.3.2a: Structure of abietic acid and reserpine showing changes in chemical shifts in Hz [recorded at 400 MHz] (*depicts the oxygen atom involved in the proposed hydrogen bond 'CO2H formed between the two compounds) Both abietic acid and reserpine are poorly soluble in water and the driving force for complex formation may be hydrophobic to exclude water. One possible interaction between the two compounds is via a hydrogen bond between the carboxylic acid group of abietic acid and a carbonyl group on reserpine, shown on figure 4.3.2a with an asterisk. The changes in proton chemical shift were used to form a theoretical model which could explain the experimental results. 4.3.3 Molecular Modelling Molecular modelling of the proposed interaction between abietic acid and reserpine was carried out by Dr. Mire Zloh at the School of Pharmacy. The NMR data were used with Macromodel software to produce a model showing the proposed interaction between the C and D rings of abietic acid and reserpine. - 173 A-COOH R-CO Figure 4.3.3a: Molecular model of abietic acid (yellow)-reserpine (green) complex. The interacting atoms detected by changes in chemical shift are represented as a CPK model, ^depicts the atoms involved in the proposed hydrogen bond formed between reserpine (R) and abietic acid (A). This work has recently been published (Smith et a l, 2005). Another possibility for a potential interaction between the two compounds is an acid- base interaction. The carbonyl group on abietic acid may deprotonate giving the proton to the tertiary nitrogen on the C-ring of reserpine. However, an acid-base interaction would raise the question of why reserpine potentiates the activity of both tetracycline and norfloxacin against TetK and NorA effluxing strains. Tetracycline has phenolic hydroxyl groups which could deprotonate and norfloxacin has an acidic carboxyl group, so there would be potential for an acid-base interaction to occur between either compound and reserpine. A study by Zloh et al. (2004) using molecular modelling has suggested that MDR inhibitors may have affinity for the substrates of efflux pumps, possibly forming complexes with them. They suggest that an inhibitor may bind the pump substrate facilitating its entry into the cell after which the complex dissociates and the drug can exert its effect. In the case of abietic - 174- acid and isopimaric acid, whether or not they are substrates for the efflux pumps, they may bind tightly to reserpine and have a slow rate of dissociation leading to a lower concentration of free drug in the bacterial cell, causing a reduction in antibacterial activity. In effect, reserpine may reduce the bioavailability of abietic acid and isopimaric acid. Zloh et a l (2004) carried out modelling of the anti-cancer drugs doxorubicin and vinblastine and showed that they may bind to rings A, B and C of reserpine. They suggest that the aromatic rings of doxorubicin interact with reserpine’s aromatic rings, whilst “the aromatic region of reserpine fitted into the curvature of the vinblastine molecule”. Norfloxacin was also shown to have affinity for reserpine’s A, B and C rings, and therefore, it is possible that for tetracycline and norfloxacin, their aromatic rings bind preferentially to the aromatic rings of reserpine, rather than hydrogen bonding being the driving force for any interaction. The results seen for isopimaric acid and abietic acid suggest that complex interactions may occur between efflux pumps, their substrates and inhibitors. This was demonstrated by Gibbons et al. (2004a) who have shown that epicatechin gallate and epigallocatechin gallate inhibit efflux of ethidium bromide from SA1199B. At low concentrations, the opposite effect was seen whereby both compounds potentiated efflux. The postulation was made that the NorA efflux pump may have high and low affinity binding sites. At low catechin concentration, only the high affinity sites would be occupied, giving a different outcome from when the low affinity sites were occupied at high catechin concentration leading to efflux. 175- 4.4 RESULTS OF EFFLUX INHIBITION ASSAYS 4.4.1 The Effect of Ferruginol on Efflux of Ethidium Bromide All of the efflux experiments, including growth of totarol mutants and transfection with the NorA efflux pump were carried out by Professor Glenn Kaatz and Susan Seo of the John D. Dingell Department of Veterans Affairs Medical Center, Detroit, Michigan, USA. Ferruginol was tested in the efflux inhibition assay which measures the effect a compound has on efflux of ethidium bromide (EtBr) from strain SA1199B which possesses the NorA pump. Figure 4.4.1 a: The effect of ferruginol on SAl 199B 100 100 80 Iw % § 60 • 40 - 20 • ' 20 0 10 20 30 40 50 Concentration (pM) In the efflux inhibition assay, 10 pM ferruginol resulted in 40% inhibition of efflux, with 50% inhibition occurring at around 17 pM . These two values correspond to 2.86 pg/ml and 4.86 pg/ml respectively. Fifty per cent inhibition of efflux occurred - 176- at the MIC value for ferruginol, and as the antibacterial activity of ferruginol would have an effect at this concentration, the experiment was stopped at 30 pM ferruginol. In the modulation assay, ferruginol at 2 pg/ml, half its MIC value, resulted in a two fold reduction in the MIC of norfloxacin against SAl 199B. Reserpine reduced the MIC of norfloxacin from 32 to 4 pg/ml but the pump was not completely inhibited, there was still some residual efflux occurring. The presence of other efflux pumps for which norfloxacin and EtBr are substrates, but which are not inhibited by reserpine or affected by ferruginol could also be responsible for some efflux of compounds (Kaatz et al., 2000). From the results, it appears that ferruginol may be a weak inhibitor of efflux, but it was not possible to determine whether it actually inhibits the pump, or has an indirect effect on the pump. If, like totarol, ferruginol has an effect on bacterial respiration, it is possible that weakening the electron transport chain would reduce the amount of energy available for the ATP-driven pump. Alternatively, if ferruginol affects membrane topology, then this could cause conformational changes in the pump which would likely lead to an effect on its efflux capabilities. 4.4.2 The Effect of Totarol on Efflux of Ethidium Bromide One of the problems with the efflux inhibition assay using ferruginol was the narrow margin between the half MIC value of 2 pg/ml and the MIC of 4 pg/ml, at which point the antibacterial activity of ferruginol would affect the assay. This would have been an even greater problem with totarol, which has an MIC of 2 pg/ml. To try to overcome this problem. Professor Kaatz decided to select for mutants with increased resistance to totarol. A knock-out mutant strain ofS. aureus, SAK1758, in which the NorA pump had been deleted was used. A stable mutant was selected which 177- exhibited an 8 -fold increase in MIC for totarol of 16 |xg/ml, compared with 2 pg/ml against SA1199B. A NorA-expressing plasmid (pK364) was then transduced into the totarol mutant, which was denoted as strain SAK3092. MIC assays confirmed that SAK3092 was expressing NorA, by an increase in norfloxacin MIC from 0.39 pg/ml to 6.25 pg/ml. The MIC for totarol was constant at 16 pg/ml, confirming that there was no efflux of totarol. Figure 4.4.2a: The effect of totarol on SAK3092 100 1 h 100 80 60 I 40 0 5 10 15 20 Concentration pM The efflux inhibition experiment showed that at 10 pM, totarol did interfere with the activity of the NorA pump, resulting in nearly 50% inhibition of EtBr efflux. 10 pM conesponds to 2.86 pg/ml, which is just under one sixth of the MIC for this totarol mulant. A higher concentration, up to 20 pM (5.72 pg/ml) did not have any incieased effect on efflux. The results of this experiment suggest that sub-inhibitory concentrations of totarol do reduce efflux by the NorA pump, although the mechanism by which this is achieved is unclear. 178- 4.5 DISCUSSION 4.5.1 Ferruginol Ferruginol is considered to be one of the diagnostic compounds of the Cupressaceae and has been found in Chamaecyparis obtusa (Fukushimaet al., 2002; Ozaki et al., 1983), Juniperus excelsa (Muhammad et al., 1992), Thuja standishii (Iwamoto et al., 2003) and Thujaplicata (Sharp et al., 2001). From other Coniferae families, ferruginol has been isolated from the Cephalotaxaceae: Cephalotaxus harringtonia (Politi et al., 2003); Podocarpaceae: Podocarpus ferrugineous (Gambie et al., 1971); Taxaceae: Torreya nucifera (Harrison and Asakawa, 1987) and Taxodiaceae: Taiwania cryptomerioides (Chang et al., 1999; He et al., 1997). Ferruginol has also been found in angiosperms and has been isolated from the roots of several species of Salvia (Lamiaceae) (Kabouche et al., 2005; Nakanishi et al., 1983; Tan et al., 2002; Ulubelen et al., 2000, 1999a, 1999b); Coleus barbatus (Lamiaceae) (Kelecom 1984) and also from Harpagophytum procumbens (Pedaliaceae), known as Devil’s Claw (Clarksonet al., 2003a). The results for ferruginol showed good activity compared with standard antibiotics against all strains, with MICs ranging from 4-16 pg/ml. This compares well with results obtained by Muhammad et al. (1992) who saw an MIC of 5 pg/ml against a penicillin-resistant strain, and Evans et al. (1999) who had 8 pg/ml against an MRS A strain. There is some contradiction between researchers’ results on ferruginol’s activity against S. aureus. Politi et al. (2003) extracted ferruginol from the seeds of - 179- Cephalotaxus harringtonia and found it active at 31.2 |ig/ml. However, Ulubelen et al. (1999b) described the activity as “slight” at greater than 250 ng/ml. Yang et al. (2 0 0 1 ), who synthesised (-)-ferruginol found that the (-) form was more active than the naturally occurring (+)-ferruginol having an MIC of 62.5 |ag/ml compared with greater than 125 pg/ml against MRS A. This discrepancy may reflect the different methods used. Ulubelen’s group used the disc diffusion assay rather than the broth dilution assay. Ferruginol, being a lipophilic compound may not diffuse well in the agar of the disc diffusion assay. Muhammad et al. (1992) and Evans et al. (1999) used the agar dilution method whereby ferruginol would be distributed throughout the agar. Yang et al, (2001) used a higher bacterial titre which may account for the higher MIC. Kabouche et al. (2005) recently reported an MIC of 16 pg/ml against strain ATCC 25923 using the disc diffusion method. The modulation assay and efflux inhibition assay results demonstrated that ferruginol had similar activity to reserpine in reducing the MIC of tetracycline against strain XU212, and gave a two-fold potentiation of norfloxacin against SA1199B. However, unlike reserpine, ferruginol also potentiated the activity of erythromycin against a strain possessing the MsrA efflux pump. The efflux inhibition assay supported these results showing that the presence of ferruginol resulted in a reduction in efflux by the NorA pump. Ferruginol showed good potentiation of oxacillin activity against an epidemic MRSA strain, and was comparable with the activity for epicatechin gallate against the same strain. An eighty-fold reduction in the MIC for oxacillin was achieved, restoring oxacillin sensitivity in a resistant strain. This potentiation of antibiotic activity by ferruginol against MRSA and effluxing strains of S. aureus has not been previously reported. Mossa et al. (2004) recently described 180- the antimycobacterial activity of ferruginol at 5 |ag/ml and its 4-fold potentiation of iisonaizid activity at half MIC against several species of mycobacteria. Ferruginol is known to have various biological activities:- antibacterial against Gram positive bacteria including Staphylococcus aureus (Muhammad et a l, 1992; Politi et œL, 2003), MRSA (Evans et al., 1999) and mycobacteria (Muhammad et al., 1992); siome antifungal activity (Chang et al., 1999); activity against human colon cancer cells (Ulubelen et al., 1999a); antihypertensive activity (Ulubelen et al., 2000); inhibition of rabbit platelet aggregation in vitro (Luo et al., 1988) and some inhibitory effect on Epstein-Barr virus early antigen activation (Iwamoto et al., 2003). Ferruginol is inactive against Gram-negative bacteria (Evans et al., 1999). Ferruginol did show very similar activity against all six S. aureus strains tested. This suggests that it targets some structure or mode of action common to all the strains, for example, it is not likely to be a (3-lactamase inhibitor, or if it is, it must have other modes of action, to account for its activity against the effluxing strains. A general mode of action for phenolic diterpenes has been suggested as cell lysis or membrane disruption, but results obtained by other researchers suggest that there may be a more selective mechanism of action (Clarkson et al., 2003a). Clarkson’s group demonstrated the antiplasmodial activity of ferruginol. They also found low cytotoxicity against mammalian Chinese Hamster Ovary cells and hepatoma HepG2 cells. This low cytotoxicity would be of value if ferruginol were to be used in a formulation as phenolic compounds are often assumed to be cytotoxic. This group also found that ferruginol (100 pg/ml) had no effect on erythrocyte -181 morphology, nor was lysis induced. As they stated, these results suggest that ferruginol exhibits selective biological activity. It is also unlikely that ferruginol acts by disrupting the bacterial cell membrane as it had no effect on erythrocytes. An important feature of ferruginol is that it has shown in vivo activity. This was demonstrated by Ulubelen et al. (2000) whereby ferruginol was administered intravenously to Wistar Albino rats. A reduction in blood pressure showed antihypertensive activity equal to the positive control. However, this does not mean that ferruginol would be active when administered orally, as intravenous injection by-passes the digestive system which could inactivate or modify the compound. These findings do appear to indicate a lack of toxicity if ferruginol could be injected directly into the bloodstream. 4.5.2 Totarol ,0 H Totarol has previously been isolated from the outer bark of Chamaecyparis nootkatensis (Constantine et al., 2001) and from the leaves of Chamaecyparis \ H formosensis (Lin, et al., 1999). It has also been found in several other species of conifer: the bark of Podocarpus totara (Evans et al., 1999) and P. nagi (Haraguchi et al., 1996; Kubo et al., 1992); the aerial parts of Thuja plicata (Sharp et a l, 2001); and the bark and leaves of Juniperus procera (Muhammad et al., 1995). The MIC for totarol of 2 pg/ml against the standard ATCC 25923 strain and EMRSA-15 are similar to results reported by Gary Evans’s group (Nicolson et al., 1999) of MICs of 3.2 pg/ml and 2 pg/ml against a methicillin-sensitive strain and an -182- MRSA strain respectively, and Kubo et al.'s (1992) results of 1.56 pg/ml and 0.78 pg/ml against a methicillin-sensitive and a resistant strain. It is interesting to note that both these research groups found that totarol was slightly more active against an MRSA strain than against a strain susceptible to methicillin. The modulation results demonstrated that at least a 32-fold reduction in the MIC of oxacillin against EMRSA-15 was achieved when incubated with totarol at 1 pg/ml (half MIC). Nicolson et a l (1999) reported a reduction in methicillin MIC from greater than 32 pg/ml to 4 pg/ml, an eight-fold reduction, when used in combination with 1 pg/ml totarol. These researchers also synthesised an analogue of totarol which had a methyl rather than an isopropyl group at position C-14 and which exhibited at least a 256-fold reduction in the MIC of methicillin. Muroi and Kubo (1996) also assayed totarol at half MIC and saw a 16-fold reduction in the MIC of methicillin against one MRSA strain, but only a 2-fold reduction against a different MRSA strain. Kubo et a l, (1992) also carried out combinatory studies, testing totarol in combination with an antibacterial phenolic compound, anacardic acid isolated from cashew nut shell oil. Anacardic acid was used at half its MIC and resulted in an 8 -fold decrease in the MIC of totarol against a methicillin-sensitive strain of S. aureus. Totarol showed moderate activity in potentiating the activity of standard antibiotics against the effluxing strains, comparable with the results seen for reserpine. However, as with ferruginol, it also potentiated the activity of erythromycin against an MsrA expressing strain. The efflux inhibition assay using a totarol mutant with increased susceptibility to the compound suggested that totarol affects the ability of - 183- the NorA pump to efflux compounds from the cell, and that this activity is likely to be separate from totarol’s antibacterial action. Totarol’s activity against effluxing strains of S. aureus has not been previously reported and publication of this work is currently in preparation (Smith et a l, 2006). Mossa et al. (2004) found totarol to be active against several species of mycobacteria (M intracellulare, M. smegmatis, M. xenopei and M chelonei) and, that at half MIC, it caused an 8 -fold potentiation of isonaizid activity against these strains. Constantine et al. (2001) also saw activity for totarol at 16 pg/ml against Mycobacterium tuberculosis. Although totarol has been shown to have activity against several Gram positive species, including Propionibacterium acnes (Kubo et al., 1992), it is inactive against Gram-negative bacteria (Evans et al., 1999) and fungi (Kubo et al., 1992). Clarkson et al. (2003b), saw antiplasmodial activity for totarol as well as ferruginol and found that its antiplasmodial activity was 40-fold greater than its cytotoxic activity. This indicates a potentially useful separation between the concentration required for activity and that at which totarol is toxic. They also synthesised totarol amino alcohol derivatives which had greater antiplasmodial activity than totarol. The Evans group (Nicolson et al., 1999) performed incorporation assays using radio labelled thymidine and A-acetylglucosamine, precursors of DNA and peptidoglycan synthesis. Totarol did not have any effect on cell protein levels and they concluded that the antibacterial action of totarol is not directed at DNA or peptidoglycan synthesis. This same group also used biotinylated ampicillin as a substrate in Western Blotting, to detect levels of PBP2’ in MRSA before and after exposure to - 184- methicillin and totarol. MRSA cells produced PBP2’even in the absence of methicillin, and addition of a totarol derivative did not affect levels of production of the protein. A difference in the level of PBP2’ expression was seen only after exposure to methicillin which resulted in increased levels of PBP2’. Addition of a sub-inhibitory concentration of the totarol derivative then resulted in a reduction in the amount of PBP2’ protein seen on the Western Blot. They concluded that potentiation of methicillin activity by totarol is by interference with de novo synthesis of PBP2’. Nicolson et al. (1999) assayed totarol for cytotoxicity against three mammalian cells and found that it was not toxic at concentrations of less than 49 p.M (14 pg/ml). These researchers also carried out in-vivo assays on totarol using a mouse model infected with methicillin-sensitive S. aureus, but totarol was ineffective against the infection. Haraguchi et al. (1996b) reported that totarol inhibits NADH-cytochrome c reductase of the respiratory electron transport chain, but they also found that totarol had potent antioxidant activity, inhibiting rat liver microsomal and mitochondrial lipid peroxidation at 4.79 and 0.47 pM respectively (1.37 and 0.13 pg/ml) (Haraguchi et al., 1997). Totarol was also shown to protect red blood cells against oxidative hemolysis (Haraguchi et al., (1996a). The Villalain group (Micol et al., 2001; Mateo et al., 2 0 0 0 ) reported that totarol was located in the inner region of the membrane bilayer and disrupted the packing of phospholipids, intercalating between the hydrocarbon chains of the phospholipids. They suggested that membrane disruption is likely to be the mode of antibacterial action of totarol and dispute Haraguchi’s postulation that its antibacterial activity may involve inhibition of respiration. - 185- pointing out that Shapiro et al. (1998) showed that under anaerobic conditions, totarol inhibited the growth of facultative and obligate anaerobic bacteria. Nicolson et al. (1999) initially agreed with Haraguchi’s observations and concluded that “the primary staphylococcal target for totarol is the respiratory chain”. However, from later experiments using rat liver mitochondria, they found that totarol had an inhibitory effect on respiration only at 50 pM and above, a concentration at least ten fold higher than its MIC for antibacterial activity. To strengthen their conclusions, they also found that 7-oxototarol, which was inactive at 32 pg/ml in the MIC assays, was slightly more effective than totarol in affecting respiration (Evans et al., 2000a). It can be seen that totarol has different activities at different concentrations and that its mode of antibacterial action has yet to be fully elucidated. Its location in the phospholipid bilayer, as described by Villalain’s group does support the view that it may function by disrupting membrane structure. Alternatively, if positioned in the membrane, totarol might be in a suitable location to interfere with PBP2’, which exerts its activity on cell wall synthesis. Totarol, may bind to PBP2’ and prevent it from binding to precursor molecules of cell wall synthesis, functioning in a similar manner to P-lactams binding to other penicillin binding proteins. However, other results, including its low toxicity to mammalian cells and antioxidative activity, suggest a more subtle mode of action than membrane disruption. Totarol’s toxicity towards Gram-positive bacteria at low concentration, but low cytotoxicity to mammalian cells, combined with antioxidative activity may be of value in the potential development of an antibacterial product. 7a-hydroxytotarol was four to eight-fold less active than totarol with MICs of 8 - 16 pg/ml. This is in agreement with results seen by Evans et al., where the compound 186- had a 16-fold higher MIC (32 gg/ml) than totarol against an MRSA strain. Studies by Pettit et al., (2004) found that 7a-hydroxytotarol had activity against six human cancer cell lines. 4.5.3 Sempervirol Sempervirol has been previously isolated from the leaves of Chamaecyparis formosensis (Lin et al., 1999) and from OH other Cupressaceae species: the resin of Cupressus sempervirens (Mangoni and Caputo, 1967) and the seed of Thujopsis dolabrata (Hasegawa and Hirose, 1981). Cambie et al., (1983) also found sempervirol in the wood of Podocarpus neriifolius. The results showed that sempervirol had good activity of 2 pg/ml against all strains, although, surprisingly, it was inactive in the modulation assays. It has previously been reported to have good antibacterial activity against Gram-positive bacteria, especially Propionibacterium acnes (Shimatani et a l, 1989). Several sempervirol- type diterpenes have recently been isolated from the leaves of C. obtusa (Kuo and Chan, 2005), some with cytotoxic activity against cancer cells. 4.5.4 Sugiol Sugiol is another compound commonly found in the Coniferae and has been isolated from the leaves of Chamaecyparis formosensis (Lin et al., 1999), the bark of Calocedrus formosana (Chao et a l, 2005) and Thuja standishii (Iwamoto et a l, 2003) of the Cupressaceae. The compound has also been extracted from the resin of Podocarpus ferrugineous (Cambie et a l, 1971) and root bark o i P. nagi (Haraguchi 187- et a l, 1997; Kubo et a l, 1992). Sugiol has been found in angiosperms in Salvia miltiorhiza (Tezukaet al., 1997) and in the roots of Aegiphila ihotzkyana (Costa- Lotufo et a l, 2004). Sugiol was inactive in the MIC assay at 10 pg/ml and was not assayed at a higher concentration due to a shortage of material, but Kubo et al. (1992) reported that it had no antimicrobial activity at 400 pg/ml. However, there are reports of antioxidant and anti-inflammatory activity for sugiol. Using rat liver microsomes, Haraguchi et al. (1997) found that at 35 pM sugiol inhibited lipid peroxidation by over 50%. Chao et al. (2005) reported that sugiol had activity in inhibiting inflammatory cytokines and was also active in suppressing the activation of mitogen-activated protein kinases (MAPKs) involved in signalling pathways. Sugiol also had antiproliferative effects on tumour cell lines and did not cause lysis of mouse erythrocytes (Costa-Lotufo et al., 2004). 4.5.5 Pisiferol, 5-Epipisiferol and Formosanoxide OH OH OH Pisiferol 5-Epipisiferol Formosanoxide Pisiferol has been found in the leaves of Chamaecyparis pisifera (Xiao et al., 2001; Yatagai and Takahashi, 1979) and C. formosensis (Lin et al., 1999). It has also been isolated from the aerial parts of Lepechinia meyeni from the Lamiaceae (Bruno et al., 1991). Several patent applications have been filed by Japanese groups for the use of - 188 - pisiferic acid and its derivatives in treatments for skin disorders including acne, dandruff, fungal infections and deodorants. Both pisiferol epimers had the same antibacterial activity of 8 - 16 pg/ml. This compares well with MICs found by other researchers of 12.5 pg/ml, (Kobayashi et al., 1988) and 25 pg/ml (Xiao et al., 2001). These results suggest that the stereochemistry of the A/B ring junction and the resultant changes in shape of the A and B rings does not affect activity. Kobayashi’s group found that pisiferol was inactive against Gram-negative bacteria (1988), but had good activity against the rice blast fungus, Pyricularia oryzae, (1987). Yatagai et al. (1994) saw good antioxidant activity for pisiferol, it being more active than a-tocopherol and it also displayed weak activity against the house dust mite Dermatophagoides pteronyssinus. 5-Epipisiferol was isolated from the cones of Sequoia sempervirens (Son et al., 2005), although this group identified the compound as 20-hydroxyferruginol. They found the compound inhibited the growth of several human tumour cell lines. Formosanoxide was inactive in both the MIC and modulation assays. It was isolated by Hsu et al. in 1995 and there is no other record of this compound in the literature or any information about its biological activity, or lack of activity. 4.5.6 /ra/is-Communic Acid ^ra«5 -Communic acid has been isolated from the leaves of Chamaecyparis formosensis (Lin et al., 1999) and the cones of C. obtusa (Fukushima et a l, 2002). It has also been COOH detected in numerous conifer species, including: the leaves - 189 o f Juniperus procera (Muhammad et al., 1995) and Thujopsis dolabrata (Takahashi et al., 2003); the roots of Araucaria angustifolia (Fonseca et al., 2000); the seeds of Sciadopitys verticillata (Hasegawa and Hirose, 1985) and the oleoresin from Chinese pine species (Song et al., 1995). tra« 5-Communic acid has also been isolated from a completely different category of the plant kingdom, the North American liverwort Porella navicularis (Bungert et al., 1998). In the MIC assays, ^ra«5 -communic acid was active against all strains at 8 - 16 pg/ml. Muhammad et al. (1995) found that (+)-Z-communic acid was two-to-three- fold more active against Gram-positive bacteria than the (+)-E form, with an MIC of 7.5 pg/ml against S. aureus compared with 15 pg/ml for the (+)-£' form. Their results also showed slight activity for (+)-Z-communic acid against E. coli at 50 pg/ml, but that the (+)-£' form was inactive at 1 0 0 pg/ml. 4.5.7 4p-Hydroxygermacra-1 (10)-5-diene 4p-Hydroxygermacra-l(10)-5-diene has been extracted from Juniperus communis (San Feliciano et al., 1995) and is also present in angiosperms, in the Myrtaceae (Cornwell et al., 2001) and the Zingiberaceae (Tchuendem et al. 1999). The compound exhibited only slight anti-staphylococcal activity (128-256 pg/ml) and did not have any modulating activity. However anti plasmodial activity against Plasmodium falciparum strains has been reported for this compound (Tchuendem et al., 1999). - 190- 4.5.8 Isopimaric Acid and Abietic Acid CO2H Isopimaric acid Abietic acid Isopimaric acid is a resin acid found in Pinus species. It has also been found in other conifer families, including the Taxaceae and has been isolated from the leaves of Torreya nucifera (Harrison and Asakawa, 1987). Soderberg et al. (1990) reported activity for both isopimaric and abietic acids against Gram-positive bacteria, including S. aureus. Reported biological activities for isopimaric acid include activation of large-conductance calcium activated potassium channels in HEK cells (Imaizumi et a l, 2002). Furthermore, studies on rainbow trout revealed that it activates calcium release from intracellular stores (Rabergh et a l, 1999) and decreases intracellular pH of hepatocytes (Nikinmaa et a l, 1999). Additionally, isopimaric acid inhibits 12-O-tetradecanoylphorbol-13-acetate induced ornithine decarboxylase, an enzyme involved in tumorigenesis (Chang et a l, 2000). This diterpene also displays feeding deterrent properties against the gypsy moth (Raffa et a l, 2002). Abietic acid is one of the main components of pine oleoresin and is found in many conifer species (San Feliciano et a l, 1993). It has also been detected in Angiosperms and was isolated from Pimenta racemosa from the Myrtaceae (Fernandez et a l, 2001). There are contrasting reports on the toxicity of abietic acid, as a major 191 - component of oleoresin, it has been reported to cause allergies such as eczema and respiratory problems in saw mill workers. It has also been described as an environmental toxin. Conversely, there are reports which suggest that abietic acid has an effect in reducing allergic responses, such as inflammation. Ulusu et al. (2 0 0 2 ) found that abietic acid inhibited 5-lipoxygenase activity, an enzyme that leads to the formation of leukotrienes which function in inflammatory and allergic responses. Abietic acid also showed anti-inflammatory activity in the rat paw odema and mouse ear oedema assay (Fernandez et al., 2001). Considerable investigation by Karlberg and co-workers has revealed that abietic acid itself is not an allergen, but that its oxidation products can cause allergic responses (Karlberg, 1989; Karlberg et al., 1985). Abietic acid is used commercially, in skin and hair formulations and treatments for bums (San Feliciano et a l, 1993). As with totarol, Villalain (1997) carried out studies on abietic acid using phosphatidylcholine model membranes. He suggested that abietic acid locates to the upper part of lipid membranes with the carboxyl group in close proximity to the phospholipid ester groups. 4.5.9 Opiopanonyl Acetate and Torulosal Oplopanonyl acetate was first isolated from Chamaecyparis pisifera by De Bruyn et al. (1990). It was inactive in the MIC and modulation assays and a literature search did not reveal any further references to this compound. Torulosal was inactive against all strains except EMRSA-16 and there is no report in the literature of antibacterial activity for this compound. Tomlosal has been isolated from the resin of a stem canker of Chamaecyparis obtusa (Yamamoto et al., 1997) and this group did see activity for tomlosal in inhibiting the growth of the fungus Sarea resinae. - 192- 4.6 STRUCTURE ACTIVITY RELATIONSHIPS The isolated compounds, shown below, were all diterpenes or sesquiterpenes with very similar structures. However, there were considerable differences in their anti- staphylococcal and modulation activities. Figure 4.6: isolated compounds CO2H trans-Communic acid Isopimaric acid Abietic acid OH OH HOH2C Pisiferol 5-Epipisiferol Formosanoxide OH OH OH Totarol Ferruginol Sempervirol OH .OH t)H 7a-Hydroxytotarol Sugiol - 193- o OH HiCL ,0 — C— CHi OH H 4p-Hydroxygermacra- Oplopanonyl acetate Torulosal l(10)-5-diene 4.6.1 Phenol and Isopropyl Groups Totarol and sempervirol and ferruginol are isomers, however in the MIC assays, totarol and sempervirol had anti-staphylococcal activity at 2 pg/ml, which was 2 to 8 - fold greater than ferruginol. The compounds only differ in the position of the phenolic hydroxyl and the isopropyl group on the C-ring. In both totarol and sempervirol, the hydroxyl is at C-13, whereas in ferruginol it is sited at C-12. Evans and co-workers (1999) suggested that in ferruginol the hydroxyl is “scissored” by the isopropyl group which accounts for its lower activity compared with totarol. If this were the case, however, it is difficult to see why this “scissoring” would not also occur in totarol. The results seen for sempervirol and totarol with the hydroxyl at position 13, but with the isopropyl group at position 12 and 14 respectively, suggest that a positional effect may account for the difference in activity, with the hydroxyl group occupying the optimum position at C-13 compared with C-12 in ferruginol. The results also indicate that the phenol group is important for antibacterial activity. Work carried out by Evans et al. (2000a; 2000b; 1999) supports this view, as although they synthesised numerous totarol analogues, they did not find one which 194- had greater antibacterial activity than totarol. Evans’ group also found that replacement of the C-13 phenolic OH group and C-14 isopropyl of totarol with hydrogen atoms resulted in no activity at 32 pg/ml. 4.6.2 Lipophilicity and SARs Studies carried out by Lien et al. in 1968 suggested that there is a correlation between the lipophilic character of a molecule and its antibacterial activity which could be expressed by its 1-octanol-water partition coefficient log P. Hydrophilic compounds would not be able to easily cross a lipid membrane and would likely remain in the first aqueous phase encountered. Strongly lipophilic compounds might also get trapped by binding tightly to lipid phases. This could be a particular problem for highly active compounds which would be used at low concentrations. Strong binding to lipids would deplete the level of the compound resulting in a reduction in bioavailability of the fi'ee drug. Lien’s conclusion (1968) was that each compound will have an ideal partition coefficient at which it can move freely in a lipid cell wall or membrane and reach its cell target, and suggested that the ideal lipophilic character (log Po) for Gram-positive bacteria was 6 . Many phenolic diterpenes are thought to be active against Gram-positive bacteria as they can cross the cell wall due to their lipophilic nature. However, Gram-negative bacteria have a thicker cell wall with a higher lipid content than Gram-positives and the diterpenes are thought to become immobilised in the lipopolysaccharide cell wall. Totarol, ferruginol, pisiferol and related compounds are amphipathic, they are lipophilic due to the presence of the isopropyl group on ring-C and the methyls and methylenes on the A and B rings. However, the phenol group imparts a hydrophilic - 195 element to the compounds and it is these two factors which are thought to account for the ability of phenolics to cross lipid membranes and reach their target. Efflux pumps are transmembrane spanning proteins which often have several helices which span the membrane. Membrane proteins may allow passage of charged or hydrophilic molecules. It is known that membrane channel proteins adopt a conformation such that hydrophobic amino acids are on the outside of the protein in the lipid bilayer and hydrophilic amino acids, with hydroxyl, sulphydryl or carboxylic moieties face the channel and are often found at the active site of a protein, for example a binding pocket, where they may interact with the substrate. Comparison of the results for ferruginol with pisiferol, which differs only in a hydroxymethylene at C-20 rather than a methyl group, shows that ferruginol was twice as active as pisiferol. The presence of an additional hydroxyl group would increase the hydrophilicity of pisiferol which might reduce its ability to cross the bacterial cell membrane. A similar situation was seen with totarol, which had a 4-to-8-fold greater activity than 7a-hydroxytotarol. The two compounds differ only by a hydroxyl group rather than a hydrogen in the a position on C-7. Again, the presence of an additional hydroxyl group would increase the hydrophilicity of 7a-hydroxytotarol. Another possibility is that as suggested by Evans et al. (2000a), where steric crowding might occur. The hydroxyl, being larger than a hydrogen atom, might cause compression of the isopropyl group which could impact on the phenol group and possibly affect activity. 196- Increased lipophilic properties may also account for trans-commumc acid having a two-to-eight-fold greater activity than isopimaric acid or abietic acid. The carboxylic acid group at C-18 on all three compounds would impart a hydrophilic property to the molecules. In the case of ^ra«5-communic acid, the hydrocarbon chain attached to C-9 would result in a greater lipophilic character than seen in either isopimaric or abietic acid. The results suggest that both the position of a functional group and the presence of lipophilic functional groups are important factors in determining the antibacterial activity of a phenolic compound. These findings have also been reported by Shapiro and Guggenheim (1998), whose QSAR studies on phenolic compounds led them to conclude that a phenolic hydroxyl is necessary for high antibacterial activity. They also determined that hydrophobicity is an important factor and found that the presence of additional polar groups leads to a decrease in activity. Their conclusions were that hydrophobic interactions are of greater importance than steric factors or hydrogen bonding in determining antibacterial activity of phenolic compounds. 4.6.3 Oxidation State of Functional Groups Experiments carried out by Kobayashi (1988) on pisiferic acid derivatives showed that oxidation of the substituent at C-20 had a detrimental effect on activity against S. aureus. Pisiferol had greater activity than pisiferal, which in turn was more active than pisiferic acid (12.5, 25, 50 pg/ml respectively). They also found that etheration of the C-12 hydroxyl of pisiferic acid increased activity with the optimum alkyl chain length being four or five carbons. There was a dramatic fall in activity when the chain length was increased to six carbons. Acylation of the hydroxyl also increased - 197 activity, again when attached to a butanoyl chain. Modification experiments on totarol by Evans et al. (1999) did not find analogues with greater activity, however, their modifications of the C-13 hydroxyl of totarol did not involve conversion to an ether or ester with an attached hydrocarbon chain. Interestingly, the oxidation state of the C-20 group of pisiferol was reversed for miticidal activity, with pisiferic acid being the most active compound, followed by pisiferal, then pisiferol (Kobayashi et at., 1988). Studies on pimaranes from Calceolaria pinifolia by Woldemichael et al. (2003) revealed activity as low as 2 pg/ml against MRSA strains. Isopimaric and abietic acids have a carboxyl group at C-18; however, their findings suggested that a C-18 oxymethylene group in pimaranes may contribute towards higher antibacterial activity rather than a carboxylic group in the same position. This conclusion was also reached by Gigante et al. (2002) who studied dehydroabietic acid derivatives and found that an alcohol or aldehyde group at C-18 resulted in higher activity than a carboxylic acid or methyl ester at the same position. Kubo et al. (1995) carried out studies on long chain alcohols and phenols and found that the optimum chain length was seen in 1 -dodecanol with 12 carbons, and that the presence of an additional CH 2 resulted in a loss of activity. With the antibacterial compound anacardic acid, they found that a Cm side chain was most active against S. aureus, but against P. acnes a C 12 chain was more active. This led them to modify Lien’s findings, by suggesting that the log Pq value for optimum lipophilic character will not be 6 for all Gram-positive bacteria, but will differ between species. A comparison between the structures of ferruginol and sugiol shows that they differ only hy an oxygen atom rather than two protons attached to C-7. However, this is sufficient to remove antibacterial activity, as sugiol has been found to be inactive at - 198- 400 |ig/ml. The double-bonded oxygen at C-7 introduces an sp^ centre to the B-ring of sugiol and, as stated by Evans et al. (2000), this should lead to flattening of the B ring. These authors suggested that in 7-oxototarol this would cause a change in steric compression compared with the H 2 -7 methylene and H-14 methine of totarol, perhaps altering the spatial arrangement of the C-13 isopropyl and C-12 hydroxyl groups. Their results showed that 7-oxototarol was inactive against S. aureus at 32 pg/ml, but that other totarol analogues with an sp^ centre on the B-ring were active, but not as active as totarol. 4.6.4 Pisiferol Epimers and Formosanoxide OH OH OH Pisiferol 5-Epipisiferol Formosanoxide Pisiferol and 5-epipisiferol had the same activity in the MIC assays, therefore the difference in stereochemistry at the A/B ring junction did not appear to have an effect on antibacterial activity. Gigante’s (2002) investigations on dehydroabietic acid derivatives also found that the stereochemistry of the A/B ring did not affect activity. There are potentially 4 isomers at the A/B ring junction in positions 5 and 10, trans- lOp and 5a or cw-lOp and 5P as seen in 5-epipisiferol and pisiferol respectively or trans-XOa and 5P or c/ 5 -10a and 5a. For diterpenes, which isomer is formed probably depends on how the molecule is folded by the geranylgeranyl diphosphate -199- synthase. It was noticeable from using a ball and stick model, that a trans configuration gave a more planar molecule whereas the cis configuration twisted the shape of the A or B rings, but this change in conformation did not affect antibacterial activity. When the C-20 hydroxyl is changed to an ether bridge connection with C-7, as seen in formosanoxide, anti-staphylococcal activity was abolished, despite the presence of the hydroxyl and isopropyl groups on the C-ring. Formosanoxide was highly volatile in comparison with the other phenolic diterpenes and the loss of the C-20 hydroxyl would increase the lipophilicity of the compound. This increase in lipophilic character might result in formosanoxide becoming trapped in the lipid membrane and could account for the loss of antibacterial activity. Using a ball and stick model, the ether bridge reduced the flexibility of the B-ring giving a rigid centre to the molecule. Figure 4.6.4a: Ball and stick model of formosanoxide -200- 4.6.5 Potentiation of Antibiotic Activity - SARs In the modulation assays, ferruginol and totarol potentiated oxacillin activity against strain EMRSA-15 and both compounds potentiated antibiotic activity against effluxing strains, whereas sempervirol had no activity in these assays. These results suggest that modulatory activity may involve a different mode of action from antibacterial activity and that the conformation of the molecule could be important. It is possible that ferruginol and totarol exert their activity by binding or fitting into a binding pocket on the efflux pump or target molecule. A bulky isopropyl group at C- 12 may give sempervirol the wrong shape for interaction with its target. The argument for different modes of action between antibacterial and modulatory activity is reinforced by results reported by Evans and co-workers (2000a; 1999). Although this group did not find a totarol analogue with more antibacterial activity than totarol, they did synthesise an analogue with much greater modulatory activity than they saw for totarol (Nicolson et al., 1999). This analogue (416) reduced the MIC of methicillin against an MRS A strain, from greater than 32 pg/ml to less than 0.125 pg/ml, compared with 4 pg/ml when combined with totarol. The compound had a methyl group at C-14 rather than an isopropyl which would result in a less bulky molecule. However, sempervirol, which has only a hydrogen at C-14 and therefore has less steric crowding than either totarol molecule in this region, had no modulating activity, a factor which suggests the isopropyl at position C-12 is unfavourable for activity. Another totarol analogue (390) differed from totarol only by the presence of a hydroxyl at position C-6, on the B-ring, which would increase the hydrophilicity of the molecule. This analogue was also more active than totarol, reducing the MIC of methicillin to less than 2 pg/ml. -201 - .OH .OH OH Figure 4.6.5a: Totarol analogue 416 Figure 4.6.5b: Totarol analogue 390 (from: Nicolson et a l, 1999) Pisiferol, which has an additional hydroxyl compared with ferruginol, was inactive in the modulation assays and therefore an increase in hydrophilicity of the compound resulted in no modulatory activity. Although, 5-epipisiferol did cause a 4-fold potentiation of tetracycline activity against XU212, the same value seen for totarol and ferruginol. Figure 4.6.5c: R = H epicatechin gallate R = OH epigallocatechin gallate OH Epicatechin gallate (ECG) was found to reduce the MIC of oxacillin against MRSA by 250- to 500-fold (Shiota et al., 1999). Epigallocatechin gallate (EGCG), which differs from ECG only by an extra hydroxyl group on the B-ring, had much lower activity, resulting in a 4- to 64-fold reduction in oxacillin MIC. It is interesting that an extra phenolic hydroxyl on EGCG should have such an effect on activity. -202 suggesting that either the position of the hydroxyl groups is important or that the increase in hydrophilicity afforded by the extra hydroxyl was sufficient to affect the activity. Studies by Hashimoto et al. (1999) on the interaction of tea catechins with lipid bilayers, found that EGCG had less affinity for the bilayers than ECG and concluded that the additional hydroxyl on the B-ring lowered its hydrophobicity. MIC assays on ECG and EGCG (Gibbons et al., 2004a) showed that EGCG was 2 to 16-fold more active than ECG, which again suggests that potentiation of antibiotic activity is by a different mechanism from antibacterial activity. The two catechin gallate molecules are much larger than totarol and ferruginol and therefore it is unlikely that the size of the compounds is an important factor for activity, but rather, as reported by Kubo et al. (1995), the balance between lipophilic and hydrophilic functional groups of the molecule is the determining factor. The results for ferruginol and totarol showed excellent potentiation of oxacillin activity against EMRSA-15 by as much as 80-fold. However, against the effluxing strains, the potentiation of antibiotic activity was modest in comparison. This was also seen by Gibbons et al. (2004a), where ECG used at one-third MIC only resulted in a four-fold potentiation of norfloxacin activity against SA1199B. One possibility for the lower activity of these compounds against effluxing strains is that they are substrates for the efflux pumps. In the case of totarol, this was shown to be unlikely, as Glenn Kaatz (Smith et al., 2006) found that a mutant with increased resistance to totarol had the same MIC of 16 pg/ml against SA1199B as it did against a Nor A deletion mutant. If totarol were a substrate for this efflux pump, the MIC should have been much higher against SAl 199B than for the deletion mutant. Shiota et al. (1999) found that ECG had no effect on oxacillin activity against a methicillin-sensitive strain, which led them to deduce that ECG may act on PBP2’, -203 - perhaps binding to the protein or interfering with its expression. This conclusion was also reached by Nicolson et al. (1999) who suggested that modulators of methicillin activity act by inhibition of de novo synthesis of PBP2’. This suggests that if compounds which act to potentiate methicillin or oxacillin activity against MRSA, do so by interference with PBP2’, then their mode of action in potentiating antibiotic activity against effluxing strains must be different. It is possible that the modest increases in antibiotic activity against effluxing strains were due to an additive effect as the compounds are antibacterial, although used at sub-inhibitory concentrations. The efflux inhibition experiments on totarol and ferruginol however, do indicate that these compounds are weak inhibitors of efflux. Other compounds isolated from plants which have shown modulating activity include camosic acid, from Rosmarinus officinalis (Oluwatuyi et al., 2004). When used at one-third MIC (10 pg/ml), an 8-fold potentiation of erythromycin activity against the erythromycin-resistant strain RN4220 was seen. OH Figure 4.6.5d: Camosic acid HO. HOOC Anacardic acid from the cashew Anacardium occidentale has shown modulating activity when used at half MIC in combination totarol against S. aureus (Kubo et al., 1992) and also with (3-caryophyllene against P. acnes. -204 COOH HO Figure 4.6.5e: Anacardic acid Stermitz et al. (2000) isolated the flavonolignan 5’-methoxyhydnocarpin from Berberis fremontii. This compound potentiated the activity of the plant antimicrobial berberine against S. aureus and inhibited efflux of EtBr in an efflux inhibition assay. .OH OMe ‘OMe HO. OH OH Figure 4.6.5f: 5’-methoxyhydnocarpin Reserpine, which is found in Rauwofia species from the Apocynaceae, is a well known efflux pump inhibitor, potentiating antibacterial activity of some antibiotics against effluxing strains (Gibbons and Udo, 2000; Markham et a/., 1999). -205 MeO' MeO OMe o OMe Figure 4,6.5g: Reserpine OMe OMe There appears to be considerable differences between these compounds, in terms of size and functional groups. Ferruginol, totarol and anacardic acid are much smaller molecules than reserpine and 5’-metboxybydnocarpin. All of the compounds except reserpine have at least one phenolic hydroxyl. Reserpine has a secondary and tertiary amine, whereas none of the other modulators contain nitrogen atoms. All of the compounds have at least one aromatic ring, but camosic acid is the only molecule with a long unsaturated hydrocarbon chain. The factor in common to all these compounds however, is a high degree of lipophilicity, combined with a hydrophilic element to the molecules. From the MIC, modulation and efflux inhibition results seen in this project, and the results reported by many other groups, it can be seen that complex interactions are likely to be occurring between antibacterial compounds and modulators and their targets. Subtle changes in stmcture, such as the presence or absence of an additional hydroxyl group, or the position of such groups appear to have a considerable effect on activity. The addition of an extra methylene group to a hydrocarbon chain can 206- abolish antibacterial or modulating activity (Shibata, et al., 2005; Kubo et al., 1995; Kobayashi et al., 1988). The antagonistic effect of reserpine on isopimaric and abietic acids, together with the finding by Gibbons et al. (2004a) that at low concentration, epicatechin gallate potentiated efflux of EtBr by NorA, but that at higher concentrations it acted to inhibit efflux, point towards activity for these phenolic compounds which is subtle, complex and likely to involve modes of action, other than merely membrane disruption. -207 5.0 PEDICULICIDE ASSAYS AGAINST HUMAN LICE 5.1 INTRODUCTION Research on several species of Chamaecyparis has shown that extracts and compounds isolated from these species have insecticidal activity, including termiticidal, acaricidal, juvenilising and antifeedant activity. Termiticidal activity has been reported for the heartwood of C lawsoniana (McDaniel, 1989), and C. nootkatensis (Lin and Chang, 1999) against the Formosan termite Coptotermes formosanus. Extracts of C nootkatensis showed acaricidal activity against nymphs of the deer tick Ixodes scapularis (Panella et aL, 1997; 2005) and leaf extracts of C obtusa were acaricidal against the house dust mite, Dermatophagoides species (Jang et al., 2005). Juvenilising activity against the yellow mealworm beetle Tenebrio molitor was seen with seed extracts of C. lawsoniana (Jacobson et al., 1975) and in an insect antifeedant assay, diterpenes from the cones of C. obtusa (Fukushimaet al., 2002) were active against larvae of the common cutworm Sodoptera litura. In view of the insecticidal activity reported in the literature for extracts from Chamaecyparis species, it was decided to test Chamaecyparis extracts and isolated compounds against human lice with the aim of identifying actives which could potentially be used in formulations to treat head lice. It would also be interesting to determine whether isolated compounds with antibacterial activity also demonstrated pediculicidal activity. The head louse is a blood-sucking ectoparasite of humans. There are three species of louse affecting humans: the head louse, Pediculus humanus capitis, the body louse, Pediculus humanus corporis and the pubic louse Pthirus pubis. Head lice are - 208 - confined to the host’s head and do not migrate to other parts of the body. There has been considerable debate about whether head lice and body lice are sub-species of P. humanus or whether they are two distinct species (Burgess, 2004). A recent study (Leo et a l, 2005) which investigated hosts infected with both head and body lice identified, by DNA analysis, that the head and body lice on each host formed genetically distinct populations. They also showed migration of lice from head to head and body to body, but not from head to body and vice versa. However, Burgess (2004) reported that migration of lice from a heavily infested head to clothing on the upper body has been noted on several occasions. Figure 5.1: Adult head louse (from: WWW.micrographia.com) Head lice are a common problem in the UK, particularly amongst schoolchildren where they are easily passed from head to head. A recent random survey of primary schoolchildren in Wales found that more than one child in ten was infected (Roberts and Burgess, 2005). Body lice live on clothing and are comparatively rare in wealthy countries where they are mainly found on homeless persons who do not have a change of clothing. Body lice can also be a problem in poor countries, in areas -209- with unsanitary conditions. Head lice will survive for about one day away from the host, whereas body lice can survive for longer periods away from the body. Head lice are considered to be merely a social problem, however, body lice can transmit the agents of serious diseases including typhus, trench fever and epidemic relapsing fever, which can cause epidemics in poorly developed countries in areas of overcrowding (Heukelbach and Feldmeier, 2004). 5.1.1 Head Lice Treatments There are two major problems associated with current treatments for head lice. The first is the emergence of resistance to insecticide treatments, and the second concerns the safety of such treatments; many parents are reluctant to use an insecticide on their children’s hair and scalp. One cause of resistance is thought to be due to the over use and misuse of pediculicides (Burkhart, 2004; Burgess, 2003), as they are often used as a preventative measure when lice are not present. At the time of writing (October 2005), there is a trial taking place in pharmacies in the North of England, whereby the pharmacist checks for the presence of head lice before selling the treatment to prevent unnecessary use (Christine Brown, Insect R&D, personal communication). Another problem is the incorrect use of treatments, for example not applying enough of the product or not repeating treatment if instructed (Burgess, 2003). Most head lice treatments do not kill the eggs and therefore have to be reapplied about a week after the first application to kill any newly hatched lice. ‘Nits’ are the eggshells left after the lice have hatched. The standard treatment for head lice is an insecticide in the form of a suspension or solution. Products containing pyrethrum and synthetic pyrethroids such as permethrin and phenothrin are used in many countries (Roberts and Burgess, 2005). -2 1 0 - These compounds target the membrane-bound sodium-gated channel of nerve cells, causing paralysis and death of the louse. Another commonly used insecticide is malathion, an organophosphate which inhibits acetylcholinesterase, also leading to paralysis and death of the louse. However, resistance to both malathion and pyrethroids has been seen (Hunter and Barker, 2003; Downs et a i, 1999). Methods of resistance are thought to include modification of the target binding site, increased detoxification of the insecticide and knockdown resistance kdr due to mutations in the sodium channel gene (Hill, et a l, 2005; Downs et a l, 1999). Lindane is an organochloride neurotoxin, prescribed as a pediculicide in the United States, but due to its transdermal absorption, the Food and Drug Administration has expressed concerns about its safety (Burkhart, 2004). Figure 5.1.1a: Head louse egg Figure 5.1.1b: Hatched egg (nit) (from: www.micrographia.com) Many parents are using the ‘combing’ method on wet or dry hair, or essential oil based and ‘natural’ remedies to try to treat head lice infestations (Craddock and Wright, 2004). However there is debate over whether these methods are effective or -211 - indeed safe, as they have not been tested for toxicity and essential oils can cause irritation (Burgess, 1998). In the United Kingdom, the most widely used pediculicide to treat head lice is a formulation containing 0.5% (/-phenothrin (Burgess et a l, 2005), which is used in Full Marks Liquid® manufactured by SSL International. Another insecticide used is permethrin, which is the active ingredient (1%) in Lyclear Creme Rinse, produced by Pfizer. Malathion is used at 0.5% in Derbac M Liquid and Suleo M Lotion; both manufactured by Simco Products Ltd. Suleo M Lotion also contains the monoterpenes limonene and terpineol. Terpineol was active in the pediculicide assays carried out by Caroline Priestley (2002) and it has been suggested that the activity of Suleo M may be partly due to the monoterpene content (Burgess, 2003). Carbaryl is only available in the UK on prescription and its use has been questioned due to possible mutagenic activity (Downs et al., 1999). ‘Nice ’n Clear’ head lice lotion, made by Nelson’s, is an essential oil based product. This treatment contains neem seed oil as a ‘traditional natural insecticide’. Morsy et al. (2000) showed that neem oil has pediculicidal activity. 'Nice 'n Clear' also contains tea tree oil and lavender oil, and of the essential oils tested by Caroline Priestley (2002), tea tree oil was the most effective, followed by lavender oil. One of the main constituents of tea tree oil is terpinen-4-ol and (+)- and (-)- terpinen-4-ol were the most active monoterpenes identified in tests by Priestley. Lavender oil contains linalool which also has pediculicidal activity, therefore, these monoterpenes may contribute to the activity of the product. Other methods to remove head lice include wet combing or application of a hair conditioner followed by wet combing, a technique described as ‘bug busting’, the rationale being that the conditioner coats the hair shaft and prevents the lice from -212 clinging to the hair. A recent study (Hill et a l, 2005) claims a 57% cure rate for the ‘Bug Buster Kit’ compared with only a 13% cure rate for proprietary pediculicides, although the methodology used in this study has been questioned (Dawes, 2005). Another method is dry combing and removal of nits with the fingers. These methods are not thought to be effective in the long term (Roberts and Burgess, 2005; Kmietowicz, 2003). A new treatment has recently been published by Burgess et al. (2005) whereby a 4% dimeticone lotion was shown to be effective in a randomised controlled equivalence trial. The results showed a 70% cure for the group treated with 4% dimeticone compared with 75% cure for the group treated with phenothrin (Full Marks Liquid®). Their in vitro studies showed that within five minutes of application, the lice were irreversibly immobilised. Dimeticone is not a neurotoxic insecticide, it acts by coating the lice, rendering them unable to manage internal water and, unlike some insecticides, dimeticone is not absorbed through the skin. Ivermectin is used to treat nematode infections and although it is not prescribed for pediculicidal use it has been used both topically and orally to treat head lice (Burkhart, 2004; Youssef et a l, 1995). Ivermectin acts by increasing the permeability of nerve and muscle cells to chloride ions leading to paralysis and death of the louse. 5.1.2 The Orlando Strain of Clothing Lice Head lice (Pediculus capitis) are highly adapted to their host environment. They use their claws to cling to the host’s hair a short distance from the scalp which creates a 213- humid environment. They take regular feeds on the host’s blood. Head lice do not colonize hair on other parts of the human body and do not colonize other animals. They will not survive for long away from the host, being very fragile and susceptible to dehydration. This makes them unsuitable for in-vitro pediculicide assays as false positives could easily occur. Clothing lice are more robust and can survive for a time away from the host when clothing is removed. They feed on human blood several times a day, which again is problematic for in-vitro assays. To overcome these problems, a strain of Pediculus humanus corporis was bred by Culpepper (1948) in the 1940’s. Described as the Orlando strain, the lice were accustomed to feed on rabbit blood rather than human blood and to have one daily feed only, as opposed to several small feeds per day. -214- 5.2 METHODS The lice assays were carried out at Insect R & D Limited in Fulboum Cambridge, and later at tbeir new site in Sbepretb, Hertfordshire. 5.2.1 Filter Paper Disc Assay Monoterpenes and essential oils have been tested for pediculicide activity using the filter paper method (Priestley, 2002; Morsy, et al., 2000). This involves adding the sample to a filter paper disc held in a glass petri dish and allowing the sample to spread out and fully saturate the paper. The paper is bung to dry in a fume hood and returned to the petri dish once dry. Lice are placed on the filter paper and the petri dish covered with a glass lid and placed in an incubator. The lice will be in direct contact with the test sample via tbeir legs, but volatile monoterpenes and essential oils will also evaporate creating vapour which would then be in contact with the whole louse and could possibly penetrate the louse via the cuticle or some other route. However, when this method was attempted for crude extracts and VLC fractions it was found that the samples did not spread evenly over the filter paper. Some of the samples were coloured and light and dark patches of colour could be seen on the filter paper indicating an uneven distribution of sample. It was also expected that the samples would contain less volatile larger molecules such as diterpenes which would not vaporise as readily as monoterpenes and therefore, the lice might only be in contact with the sample via their legs. This factor was highlighted by Burkhart and Burkhart (2001) who suggested that the paper disc assay favours volatile compounds which could be absorbed through the spiracles. The paper disc assay was therefore deemed unsatisfactory since the route of entry and -215 mode of action of any actives would be unknown, and it was considered essential that the whole louse should be in contact with the test sample. It was decided therefore to follow the dip method currently used at Insect R&D, whereby the lice are held on a piece of gauze and dipped into the test solution, ensuring that the louse is completely immersed in the sample. 5.2.2 Dip Method Pediculicide Assay Between 18 and 23 lice were used in each assay. The lice were carefully placed onto a piece of fine meshed gauze 1.5 cm^ which the lice cling to and the gauze was placed in a small petri dish (5 cm diameter). Using forceps, the gauze with the lice attached was dipped in the test solution for 1 0 seconds, then blotted on paper tissue and returned to the petri dish. The lice were left in an incubator at 30°C and 70% relative humidity for one hour. All samples were dissolved in propan- 2 -ol/H20 70:30. Crude extracts and hexane vacuum liquid chromatography (VLC) fractions were assayed at 5% and pure compounds at 1%. Propan- 2 -ol/H20 70:30 was used as a control. After incubation, the lice were washed in Stiefel shampoo base/warm water 1:15. The diluted shampoo was added to the petri dish which was gently shaken. The lice and gauze were then tipped out into a small tea strainer and rinsed with warm water, blotted on a paper tissue and placed in a new petri dish. The lice were returned to the incubator and left overnight. The following morning, the lice were scored by Maxine Harris at Insect R&D. The number of dead, morbid and alive lice was scored for each test sample and the percentage mortality calculated. Lice classified as ‘morbid’ were not actively moving around, but could be moving their antennae, head, gut or -216- legs. Lice have the ability to reach an apparently morbid state, but may recover just a few hours later (Burkhart and Burkhart, 2001). For this reason the lice were not scored at regular time intervals but left undisturbed in the incubator overnight. Once a louse has reached a morbid state after an overnight incubation, it is unlikely to recover and the figures for both dead and morbid lice were included in the percentage mortality figures. The Abbot’s Correction formula (Abbot, 1925) was applied to the results to take into account mortality in the control. Abbot’s Correction Formula: % Mt - %Mc M = ______X 100 1 0 0 -% Me M = Corrected percentage mortality Mt = Mortality in tested group Me = Mortality in control group The aim of the dip method assay was to try to replicate as far as possible the conditions under which a head lice treatment would be used by the consumer. The lice were placed on fine gauze which they cling to as a substitute for human hair. They were dipped in the test solution for only ten seconds, then blotted dry, this was done to reflect what happens when a treatment is applied to wet hair. It will get diluted and will not be in contact with the lice at its original concentration. If applied to dry hair the fluid will still disperse over the hair shaft surface and will also start to evaporate (Burgess, 2004). After one hour the lice were washed in diluted shampoo base to take account of the fact that some treatments are washed out. The lice were kept in an incubator at 30°C and 70% relative humidity to replicate as closely as possible the environmental conditions the lice would experience close to the human scalp. -217 5.3 RESULTS The crude extracts and hexane VLC fractions of Chamaecyparis lawsoniana and C. nootkatensis were assayed in the lice screen at 5%. Unfortunately, there was a high mortality rate of 40% in the control group (Table 5.3.1). After the Abbot’s Correction formula had been applied to the results, it could be seen that some samples resulted in considerably less mortality than the control, for example C. nootkatensis acetone extract -58.7% mortality. However, even after taking account of the control mortality, some samples had a mortality rate of over 50%. The crude CHCI 3 extract from both C. lawsoniana and C. nootkatensis had a mortality rate of 68.3% and 58.3% respectively. VLC fraction 5 of C. lawsoniana resulted in 71% mortality and VLC fraction 6 o f C. nootkatensis 83.3% mortality. The C lawsoniana VLC fraction 2 had a mortality rate of 66.7%. This sample was from VLC carried out on a crude hexane extract and the second fraction was considered likely to contain volatile monoterpenes and oils. Essential oils and monoterpenes have been extensively tested against lice (Lahlou, 2004; Priestley, 2002; Morsy, et al., 2000) and therefore, as a dereplication process, GC-MS was carried out on a sample of VLC 2 to try to identify the main constituents of this fraction. The GC-MS results showed the presence of low molecular weight compounds with m/z of 136 and 204. The NIST compound library identified the major components as likely to be limonene, a-pinene, naphthalene, P-pinene, caryophyllene and thujopsene. Work carried out by Priestley (2002) has shown that low molecular weight mono-oxygenated monocyclic terpenoids have high pediculicidal activity and -218 Table 5.3.1 : Results of initial screening of Chamaecyparis extracts Sample Total Dead Morbid Alive Mortality Abbot’s % correction Chamaecyparis lawsoniana Hexane 2 0 1 0 3 7 65.0 41.7% CHCI3 2 1 1 2 5 4 81.0 68.3% Acetone 2 0 11 1 8 60.0 33.3% MeOH 21 7 1 13 38.1 3.2% VLC 2 2 0 16 0 4 80.0 66.7% VLC 3 2 0 9 2 9 55.0 25.0% VLC 4 2 1 6 2 13 38.0 -3.3% VLC 5 23 16 3 4 82.6 71.0% VLC 6 18 2 3 13 27.8 -20.3% VLC 7 18 3 2 13 27.8 -20.3% VLC 8 2 0 9 1 1 0 50.0 16.7% VLC 9 19 6 1 1 2 36.8 -5.3% Chamaecyparis nootkatensis Hexane 2 1 3 2 16 23.8 -27.0% CHCI3 2 0 1 2 3 5 75.0 58.3% Acetone 2 1 1 0 2 0 4.8 -58.7% VLC 3 2 0 6 2 1 2 40.0 0 .0 % VLC 5 2 0 1 2 2 6 70.0 50.0% VLC 6 2 0 17 1 2 90.0 83.3% VLC 7 18 8 3 7 61.1 35.2% VLC 8 2 1 8 3 1 0 52.4 20.7% VLC 9 2 0 5 1 14 30.0 16.7% Control 2 0 5 3 1 2 40.0 -219- that non-oxygenated terpenoids including a-pinene, P-pinene and limonene have little or no activity. However, it is possible that collectively the terpenes present in VLC 2 have a higher pediculicidal activity than they do as individual entities. Since the main constituents of VLC fraction 2 had already been subjected to pediculicide testing by others, it was decided not to carry out further tests on this fraction. The assay was repeated on C. lawsoniana VLC 5 at 5%, C nootkatensis VLC 5 at 2.5% and 5% and VLC 6 at 0.5%, 1%, 2.5% and 5%. Table 5.3.2: Results of repeat pediculicide assay on selected VLC samples Alive Mortality A bbot’s Sample Total Dead M orbid % C orrection C.L. VLC 5 @ 5% 2 2 0 0 2 2 0 0 % C.N. VLC 5 # 5% 2 2 6 5 11 50 47.4% " @ 2.5% 2 0 7 0 13 35 31.6% C.N. VLC 6 @ 5% 2 0 6 2 1 2 40 36.8% " @ 2.5% 2 0 1 1 18 1 0 5.3% " (% 1% 2 0 9 11 0 1 0 0 1 0 0 % " # 0.5% 2 0 6 4 1 0 50 47.4% Control 2 0 1 0 19 5 C.L. = Chamaecyparis lawsoniana C.N. = Chamaecyparis nootkatensis The 5% sample of C. nootkatensis VLC 5 gave a mortality rate of 47.4%, which was the same result obtained in the previous assay. At a lower concentration (2.5%) the mortality rate fell to 31.6% which suggested that the activity might be concentration dependent. However, the results achieved for the other VLC samples were not consistent with the results from the initial screening. The C. lawsoniana VLC 5 sample was inactive compared with the previous assay where it caused 71.0% mortality. C. nootkatensis VLC 6 was the most active sample in the previous assay -2 2 0 - with a mortality rate of 83.3%. In this assay, however, the 5% sample resulted in 36.8% mortality, with the 2.5% concentration only causing 5.3% mortality. Interestingly, the 1% sample had a 100% mortality rate with the 0.5% sample resulting in 47.4% mortality. It is possible that the extract had different activity at a low concentration than at a high concentration, for example, the target may have high affinity and low affinity binding sites. Therefore, the two samples were retested, C lawsoniana VLC 5 at 5% and C. nootkatensis VLC 6 at 2.5% and 5%. Table 5.3.3: Repeat of pediculicide assay on C.L. VLC 5 and C.N. VLC 6 Alive Mortality Abbot’s Sample Total Dead Morbid % Correction C.L. VLC 5 @ 5% 20 19 1 0 100 100% C.N. VLC 6 @ 5% 20 17 0 3 85 81.5% " @ 2.5% 20 20 0 0 100 100% Control 21 3 1 17 19 Again, there was high mortality in the control sample. However, this time the C. lawsoniana VLC 5 sample showed a 100% mortality rate which compares favourably with the rate of 71% seen in the initial screening. Both concentrations of C. nootkatensis VLC 6 had high activity, although the 2.5% sample gave 100% mortality compared with 81.5% for the 5% sample. Ferruginol, isolated from C. lawsoniana and abietic acid obtained from Aldrich, were tested in the pediculicide assay. -221 - Table 5.3.4: Results for abietic acid and ferruginol in the pediculicide assay Mortality Abbot’s Sample Total Dead Morbid Alive % Correction Abietic acid 20 1 3 16 20.0 -33.3% Ferruginol 21 6 2 13 38.1 -3.2% Control 20 5 3 12 40.0 There was high mortality in the control in this assay, which after application of Abbot’s Correction formula showed both compounds to be inactive. However, even without this correction, it could be seen that neither compound was very active at 1%. Compounds isolated from C. lawsoniana and C. nootkatensis were assayed in the pediculide test at 1%. Table 5.3.5: Activity of Chamaecyparis sesquiterpenes and diterpenes against adult lice Mortality Abbot’s Sample Total Dead Morbid Alive % Correction 5-Epipisiferol 2 0 5 5 1 0 50 47.4% Oplopanonyl acetate 23 4 19 0 1 0 0 1 0 0 % Torulosal 2 1 9 8 4 81 80.0% Formosanoxide 2 0 9 11 0 1 0 0 1 0 0 % 4p-Hydroxygermacra- 19 5 14 0 1 0 0 1 0 0 % l(10)-5-diene Totarol 2 0 4 0 16 2 0 15.8% 7 a-Hydroxytotarol 2 0 9 4 7 65 63.2% Control 2 0 1 0 19 5 All the samples with the exception of totarol showed good pediculicidal activity, with three compounds exhibiting a mortality rate of 1 0 0 %. -2 2 2 - Oplopanonyl acetate and torulosal which gave a mortality rate of 100% and 80% respectively were assayed at lower concentrations. Oplopanonyl acetate was tested at 0.125%, 0.25%, 0.5% and 1.0% and torulosal at 0.125%, 0.25% and 0.5% . Formosanoxide and 4p-hydroxygermacra-l(10)-5-diene which had resulted in 100% mortality could not be tested further due to the paucity of the samples. Table 5.3.6: Pediculicide assay on different concentrations of active compounds Mortality Abbot’s Sample Total Dead Morbid Alive % Correction Oplopanonyl Ac. 1% 2 1 3 1 17 19 - 8 .0 % 0.5% 2 0 7 2 11 45 26.7% 0.25% 2 0 2 1 17 15 -13.3% “ 0.125% 19 8 0 1 2 42 22.7% Torulosal 0.5% 2 0 2 0 18 1 0 -20.0% 0.25% 20 4 1 15 25 0% 0.125% 2 0 5 0 15 25 0% C.L. VLC 5 - 5% 2 0 6 1 13 35 13.3% 2 0 2 0 18 1 0 -2 0 .0 % C.N. VLC 6-1% Control 2 0 0 15 25 5 Yet again, there was high mortality in the control (25%). Also, the results were not consistent with the previous results. In the first assay of oplopanonyl acetate, a 1% solution resulted in 1 0 0 % mortality, but in the second test, the mortality before Abbot’s Correction was 19% and the figure was negative (-8.0%) after the formula was applied. Due to lack of sample, the assay on torulosal was not repeated at 1%, but the 0.5% sample showed less activity than the lower concentrations of 0.25% and -223 - 0.125% which gave 25% mortality before Abbot’s Correction. The assay was repeated on C. lawsoniana VLC 5 at 5% which this time, yielded a mortality rate of 35% before Abbot’s Correction. This VLC fraction was assayed on four occasions and the results ranged from 0% mortality to 100% mortality. It was obvious that consistent results could not be obtained for these samples in this pediculicide assay and it was decided not to continue with testing the samples against clothing lice. Possible reasons for the failure of the assay are discussed in the next section (5.4). -224- 5.4 DISCUSSION A major problem encountered with the pediculicide assay was the high mortality among the control lice. This resulted in many samples showing activity, but once the Abbot’s Correction formula had been applied to take account of mortality in the control sample, the samples were considered inactive. The Orlando lice strain is no longer maintained at Insect R&D and the lice are sent from Dartford, Kent to Cambridgeshire by courier on the morning the assay takes place. It is therefore not possible to directly control the conditions the lice have been kept in, for example, the temperature, humidity, time of feeding. Another factor is transportation during which the lice are not kept in an incubator and this may cause stress resulting in high mortality. A shortage of sample material was another problem encountered when testing pure compounds. A 1% solution was considered the ideal starting concentration as concentrations of actives above 1 % can cause regulatory difficulties in developing a non-prescription product (Dr. David Small, Stiefel Research Laboratories, personal communication). However, 1 mg of sample only yields 100 pi of test solution at 1%. This volume is inadequate to fully submerge the gauze with the lice attached. 200 pi was found to be the minimum volume which would suffice to fully immerse the gauze and lice in the test solution. In instances where less than 200 pi of sample was available, the gauze was partially dipped in the solution for 1 0 seconds, then turned round and the other side of the gauze was dipped in the solution. It was decided that this would result in an exposure time for some lice of more than 1 0 seconds, but other lice might be exposed for less than 10 seconds. The alternative was to pour the solution over the gauze. The drawback with this method was that the lice might not -225 - be fully submerged, the solution might only touch the back of the louse, not the underside of the body or the legs. More sample was available with the VLC fractions and 500 pi to 1 ml of test solution was used to completely submerge the lice. However the results obtained for C. lawsoniana VLC 5 and C nootkatensis VLC 6 were still inconsistent. Ideally, all samples should be in an identical volume of solution which would be at least 500 pi to enable all the lice to be fully exposed to the test sample and for an identical length of time. However, for a starting concentration of 1%, this would require a minimum of 5 mg of sample which was not possible for isolated compounds due to the lack of material. Another limiting factor was the high per sample cost of the assay. This prevented replicates from being assayed, for example five replicates of C. lawsoniana VLC 5 at 5% on the same batch of lice. This would have facilitated statistical analysis of the results and identified problems with the assay. It would have been better to standardise the assay using crude extracts or VLC fractions where there was more sample available. The initial screen was only performed on one replicate to save cost and identify active extracts for further fractionation. The investigation was driven by the need to find single entity actives rather than a mixture. An active extract would be difficult to standardise in a formulation. A crude extract or VLC fraction taken from a different batch of starting material might not contain exactly the same components or ratio of components, since the production of secondary metabolites by plants is known to be affected by many factors such as the season and availability of water and nutrients. Even if all the components of an active fraction were identified, for example by GC-MS, without individual testing of each compound, it 226- would not be known which were the active constituents, making the formulation of a commercial product untenable. The results must therefore be treated as preliminary, although some of the compounds, for example oplopanonyl acetate, torulosal, formosanoxide and 4p- hydroxygermacra-1(10)-5-diene, appeared to have good pediculicide activity even under these non-ideal conditions and would be worthy of further investigation if more sample could be obtained. Studies by Yatagai and Nakatani (1994) on the activity of pisiferic acid and its congeners against the house dust mite {Dermatophagoides pteronyssinus), found that the oxidation state of the C-20 constituent affected the miticidal activity of the compounds. Pisiferic acid with a carboxylic acid group at C-20 was the most active, followed by pisiferal with a C-20 aldehyde, then pisiferol with an hydroxyl group at the same position. Therefore, the compound with the most oxidised constituent at C- 20 had the greatest activity and pisiferol, with the least degree of oxidation at C-20 was the least active. OR Figure 5.4.1: Pisiferic acid and its congeners (after Yatagai and Nakatani, 1994) Ri R2 Pisiferic acid COOH H Pisiferal CHO H Pisiferol CH 2O H H Ferruginol CH3 H O-Methyl pisiferic acid COOH CH3 Methyl pisiferate C O O C H 3 H 227- Yatagai and Nakatani (1994) also found that when the hydroxyl at C-12 was replaced with a methyoxy group, as in O-methyl pisiferic acid this had little effect on miticidal activity. However, when the C-20 carboxyl was changed to a carboxymethyl group, as in methyl pisiferate, activity was considerably reduced. They concluded that the C-12 phenol had little effect on miticidal activity and that the degree of oxidation of the C-20 substituent had a much greater impact on activity. These authors also tested ferruginol in the miticidal assay, but it resulted in only 2 0 % mortality, compared with nearly 40% for pisiferol and 80% for pisiferic acid. Interestingly, the reverse trend was seen for antibacterial activity. Kobayashi et al. (1988) found that oxidation of the C-20 substituent was detrimental to activity against S. aureus. Their results showed an order of activity of pisiferol>pisiferal>pisiferic acid. The results of the above two studies might explain why the phenolic diterpenes ferruginol and totarol, with a C-12 hydroxyl, but only a methyl at C-20 were inactive in the lice assays (-3.2% and 15.8% mortality respectively), but that 5-epipisiferol with a C-20 OH did have some activity at 47.4%. Their findings also appear to support the results seen in this project, that good antibacterial did not correlate to good pediculicidal activity. One of the main objectives of this project was to ascertain whether the most promising candidate compounds might also have dual activity as an insecticide, in addition to antibacterial properties. However, the most active pediculicidal compounds were either inactive or had only low activity in the antibacterial assays -228 against S. aureus strains. The most active anti-staphylococcal compounds were phenolic diterpenes, including totarol, ferruginol and pisiferol, but these compounds had low activity of 50% or less in the pediculicide assay. It could be tentatively suggested that anti-staphylococcal activity does not appear to correlate to pediculicidal activity, and for this and other reasons, the effects on lice were not pursued further. -229- 6.0 CONCLUSION The results from this study showed that immature conifer cones do contain active compounds against clinically relevant multidrug-resistant and methicillin-resistant strains of Staphylococcus aureus. The most active compounds were phenolic diterpenes and, although they are known compounds, the results suggested that further investigation of these active diterpenes is warranted. Ferruginol in particular potentiates oxacillin activity against MRSA strains, which has not been previously demonstrated. Ferruginol and totarol both had activity as modulators in potentiating antibiotic activity against effluxing strains, which has not been reported before. Of some interest is their activity in potentiating erythromycin activity against a strain containing the MsrA efflux pump, since no inhibitor of this pump has been reported. Results reported by other researchers, citing a range of biological activities for ferruginol and totarol suggest that these phenolic diterpenes do have some specificity in their activity, which is concentration dependent and that their mode of action is not by general disruption of the cell membrane, as is often suggested for phenolics. The results of the modelling study using reserpine and abietic acid suggested that reserpine may reduce the bioavailability of the antibacterial diterpenes abietic acid and isopimaric acid. This work support studies by Zloh et al. (2004) who suggested that complex interactions may occur between an efflux pump inhibitor and the pump substrate. -230- The efflux inhibition studies indicated that ferruginol and totarol are likely to be weak inhibitors of efflux in the SA1199B strain. The efflux inhibition experiment using a mutant strain with increased resistance to totarol separated the antibacterial activity of totarol from efflux inhibition, demonstrating that at sub-inhibitory concentrations, totarol did appear to have some activity as an efflux inhibitor. There is potential for the use of conifer compounds as antibacterials for topical application, for example in lotions or soaps. Their activity as potentiators of antibiotic activity means that they could be incorporated into a formulation in combination with an antibiotic. Conifer compounds have an advantage in that they are natural products, which are viewed favourably by the public, and they often have a pleasing fresh aroma which could encourage the use of a product, such as a handwash, which has to be used repeatedly in the clinical environment. Full structure elucidation was conducted on known compounds. This has shown discrepancies in the literature for 5-epipisiferol and rra« 5 -communic acid. The revised data for these two compounds will be published. This study will be continuing, with further investigation of Chamaecyparis nootkatensis and other conifer cones, with an emphasis on efflux inhibition experiments. It is likely that other phenolic terpenes, related to ferruginol and totarol, will have similar biological activities and may show potential as resistance modifying agents. -231 - REFERENCES Abbot, W.S. 1925. A method of computing the effectiveness of an insecticide. Journal o f Economic Entomology 18(2) 265-267. Amyes, S.G.B. 2001. Magic Bullets Lost Horizons: The Rise and Fall of Antibiotics. Taylor & Francis, London, pp 53; 99. Andersen, N.H. and Syrdal, D.D. 1970. Terpenes and sesquiterpenes of Chamaecyparis nootkatensis leaf oil. Phytochemistry 9 1325-1340. Aucken, H.M., Ganner, M., Murchan, S., Cookson, B.D., Johnson, A.P. 2002. A new strain of epidemic methicillin-resistant Staphylococcus aureus (EMRSA-17) resistant to multiple antibiotics. Journal of Antimicrobial Chemotherapy 50 171-175. Blazso, G., Gabor, M., Schonlau, F., Rohdewald, P. 2004. Pycnogenol® accelerates wound healing and reduces scar formation. Phytotherapy Research 18 579-581. Bohlmann, F., Zdero, C., King, R.M., Robinson, H. 1984. A hydroxygermacrene and other constituents from Pseudobrickellia brasiliensis. Phytochemistry 23 1798-1799. Brenwald, N.P. and Fraise, A.P. 2003. Triclosan resistance in methicillin-resistant Staphylococcus aureus (MRSA). Journal of Hospital Infection 55 141-144. -232- Bruno, M., Savona, G., Piozzi, F., De La Torre, M.C. Rodriguez, B., Marlier, M. 1991. Abietane diterpenoids from Lepechinia meyeni and Lepechinia hastata. Phytochemistry 30 2339-2343. Bungert, M., Gabier, J., Adam, K.P., Zapp, J., Becker, H. 1998. Pinguisane sesquiterpenes from the liverwort Porella navicularis. Phytochemistry 49 1079-1083. Burgess, I.F., Brown, C.M., Lee, P.N. 2005. Treatment of head louse infestation with 4% dimeticone lotion: randomised controlled equivalence trial. British Medical Journal, doi: 10.1136/bmj.38497.506481.8F (published 10 June 2005). Burgess, I.F. 2004. Human lice and their control. Annual Review of Entomology 49 457- 481. Burgess, I.F. 2003. How to advise a patient when over the counter products have failed. British MedicalJournal 326 1257. Burgess, I.F. 1998. Head lice - developing a practical approach. The Practitioner 242 126-129. Burkhart, C.G. 2004. Relationship of treatment-resistant head lice to the safety and efficacy of pediculicides. Mayo Clinic Proceedings 79(5) 661-665. -233 - Burkhart, C.N. and Burkhart, C.G. 2001. Recommendation to standardise pediculicidal and ovicidal testing for head lice (Anoplura: Pediculidae). Journal of Medical Entomology 3S(2) 127-129. Gambie, R.C., Cox, R.E., Croft, K.D., Sidwell, D. 1983. Phenolic diterpenoids of some podocarps. Phytochemistry 22 1163-1166. Gambie, R.C., Madden, R.J., Parnell, J.C. 1971. Chemistry of the Podocarpaceae. Australian Journal of Chemistry 24 217-221. Chang, H.M., Cheng, K.P., Choang, T.F., Chow, H.F., Chui, K.Y., Hon, P.M., Tan, F.W.L., Yang, Y., Zhong, Z.P., Lee, C.M., Sham, H.L., Chan, C F., Cui, Y.X., Wong, H.N.C. 1990. Structure elucidation and total synthesis of new tanshinones isolated from Salvia miltiorrhiza Bunge (Danshen). Journal of Organic Chemistry 55 3537-3543. Chang, L.C., Song, L.L., Park, E.J., Luyengi, L., Lee, K.J., Farnsworth, N.R., Pezzuto, J.M., Kinghom, A.D. 2000. Bioactive constituents of Thuja occidentalis. Journal o f Natural Products 63 1235-1238 Chang, S., Sievert, D.M., Hageman, J.C., Boulton, M.L., Tenover, F.C., Downes, F.P., Shah, S., Rudrik, J.T., Pupp, G.R., Brown, W.J., Cardo, F., Fridkin, S.K. 2003. Infection with vancomycin-resistant Staphylococcus aureus containing the vanA resistance gene. New England Journal o f Medicine 348 1342-1347. -234- Chang, S.T., Wang, S.Y., Wu, C.L., Su, Y.C., Kuo, Y.H. 1999. Antifungal compounds in the ethyl acetate soluble fraction of the extractives of Taiwania {Taiwania cryptomerioides Hayata) heartwood. Holzforschung 53 487-490. Chao, K.P., Hua, K.F., Hsu, H.Y., Su, Y.C., Chang, S.T. 2005. Anti-inflammatory activity of sugiol, a diterpene isolated from Calocedrus formosana bark. Planta Medica 71 300-305. Cheng, Y.S. and von Rudloff, E. 1970. The volatile oil of the leaves of Chamaecyparis nootkatensis. Phytochemistry 2517-2527.9 Clarkson, C., Campbell, W.E., Smith, P. 2003a. In vitro antiplasmodial activity of abietane and totarane diterpenes isolated from Harpagophytum procumbens (Devil’s Claw). Planta Medica 69 720-724. Clarkson, C., Musonda, C.C., Chibale, K., Campbell, W.E., Smith, P. 2003b. Synthesis of totarol amino alcohol derivatives and their antiplasmodial activity and cytotoxicity. Bioorganic & Medicinal Chemistry 11(20) 4417-4422. Cole, E.C., Addison, R.M., Rubino, J.R., Leese, K.E., Dulaney, P.D., Newell, M.S., Wilkins, J., Gaber, D.J., Wineinger, T., Criger, D.A. 2003. Investigation of antibiotic and antibacterial agent cross-resistance in target bacteria from homes of antibacterial product users and nonusers. Journal of Applied Microbiology 95 664-676. -235 Coll, J.C. and Bowden, B.F. 1986. The application of vacuum liquid chromatography to the separation of terpene mixtures. Journal of Natural Products 49(5) 934-936. Constantine, G.H., Karchesy, J.J., Franzblau, S.G., LaFleur, L.E. 2001. (+)-Totarol from Chamaecyaris nootkatensis and activity against Mycobacterium tuberculosis. Fitoterapia 72 572-574. Cornwell, C.P., Reddy, N., Leach, D.N. Wyllie, S.G. 2001. Germacradienols in the essential oils of the Myrtaceae. Flavour and Fragrance Journal 16 263-273. Costa-Lotufo, L.V., Silveira, E.R., Barros, M.C.P., Lima, M.A.S., de Moraes, M.E.A., de Moraes, M.O., Pessoa, C. 2004. Anti-proliferative effects of abietane diterpenes from Aegiphila ihotzkyana. Planta Medica 70 180-182. Cox, R.A., Conquest, C., Mallaghan, C., Maples, R.R. 1995. A major outbreak of methicillin-resistant Staphylococcus aureus caused by a new phage type (EMRSA-16). Journal of Hospital Infection 29 87-106. Craddock, D., and Wright, D. 2004. Parental beliefs about head lice and their management. International Journal of Pharmacy Practice 12 Supplement R42. Culpepper, G.H. 1948. Rearing and maintaining a laboratory colony of body lice on rabbits. American Journal of Tropical Medicine 28 499-504. -236 Dawes, M. 2005. Combing and combating head lice. British Medical Journal 331 362- 363. Debiaggi, M., Pagani, L., Cereda, P.M., Landini, P., Romero, E. 1988. Antiviral activity of Chamaecyparis lawsoniana extract: study with Herpes simplex virus type 2. Microbiologica 11 55-61. De Bru>’n, A., De Pooter, H.L. De Buyck, L., Jans, A.W.H., Budesinsky, M., Sedmera, P. 1990. Identification of oplopanonyl acetate isolated from Chamaecyparis pisifera. Magnetic Resonance in Chemistry 1030-1039. Dewick, P.M. 1997. Medicinal Natural Products. Chapter 5. John Wiley & Sons Ltd, Chichester, England, pp 152-184 Digrak, M., Ilçim, A., Alma, M.H. 1999. Antimicrobial activities of several parts of Pinus brutia, Juniperus oxycedrus, Abies cilicia, Cedrus libani and Pinus nigra. Phytotherapy Research 13 584-587. Downs A.M.R., Stafford, K.A., Coles, G.C. 1999. Head lice: prevalence in schoolchildren and insecticide resistance. Parasitology Today 15(1) 1-4. Eberhardt, T.L. and Young, R.A. 1996. Assessment of the anti-HIV activity of a pine cone isolate. Planta Medica 62 63-65. -237- Edwards, N. 2004. Healthcare-acquired infections: the £lbn bug. Health Service Journal 41-46. Enzell, C. 1961. The chemistry of the natural order cupressales 38: the structures of the diterpenes torulosol, torulosal and agatholic acid. Acta Chemica Scandinavica 15 1303- 1312. Evans, G.B., Fumeaux, R.H., Gainsford, G.J., Murphy, M.P. 2000a. The synthesis and antibacterial activity of totarol derivatives. Part 3: modification of ring-B. Bioorganic and Medicinal Chemistry 8 1663-1675. Evans, G.B. and Fumeaux, R.H. 2000b. The synthesis and antibacterial activity of totarol derivatives. Part 2: modifications at C-12 and 0-13. Bioorganic and Medicinal Chemistry^ 1653-1662. Evans, G.B., Fumeauz, R.H., Gravestock, M.B., Lynch, G.P., Scott, G.K. 1999. The synthesis and antibacterial activity of totarol derivatives. Part 1 : modifications of ring-C and pro-drugs. Bioorganic and Medicinal Chemistry 1 1953-1964. Fang, J.M., Hsu, K.C., Cheng, Y.S. 1989. Terpenoids from leaves of Calocedrus formosana. Phytochemistry 1% 1173-1175. 238- Fernandez, M.A., Tomos, M.P., Garcia, M.D., de las Heras, B., Villar, A.M., Saenz, M.T. 2001. Anti-inflammatory activity of abietic acid, a diterpene isolated from Pimenta racemosa var. grissea. Journal of Pharmacy and Pharmacology 53 867-872. Fisher, J.F., Meroueh, S.O., Mobashery, S. 2005. Bacterial resistance to P-lactam antibiotics: compelling opportunism, compelling opportunity. Chemical Reviews 105(2) 395-423. Fukushima, J.I., Yatagai, M., Ohira, T. 2002. Abietane-type and labdane-type diterpenoids from the cones of Chamaecyparis obtusa. Journal of Wood Science. 48 326-330. Gibbons, S., Moser, E., Kaatz, G.W. 2004a. Catechin gallates inhibit multidrug resistance (MDR) in Staphylococcus aureus. Planta Medica 70 1240-1242. Gibbons, S. 2004b. Anti-staphylococcal plant natural products. Natural Product Reports 21 263-277. Gibbons S, Oluwatuyi M, Veitch NC, Gray A.I. 2003. Bacterial resistance modifying agents from Lycopus europaeus. Phytochemistry 83-87. 62 Gibbons S. and Udo E.E. 2000. The effect of reserpine, a modulator of multidrug efflux pumps, on the in vitro activity of tetracycline against clinical isolates of methicillin -239- resistant Staphylococcus aureus (MRSA) possessing the determinant. Phytotherapy Research 14 139-140. Gibbons, S. and Gray, A.I. 1998. Methods in Biotechnology, Vol. 4: Natural Products Isolation, ed. Cannell, R.J.P. Humana Press Inc., Totowa, N.J. 209-245. Gigante, B., Silva, A.M., Marcelo-Curto, M.J., Savluschinske-Feio, S., Roseiro, J., Reis, L.V. 2002. Structural effects on the bioactivity of dehydroabietic acid derivatives. Planta Medica 68 680-684. Gilbert P. and McBain, A. 2004. Live and let die. Microbiology Today 31 62-63. Gillet, Y., Issartel, B., Vanhems, P., Foumet, J.C., Lina, G., Bes, M., Wandensech, F., Piémont Y., Brousse, N., Floret, D., Etienne, J. 2002. Association between Staphylococcus aureus strains carrying gene for Panton-Valentine leukocidin and highly necrotising pneumonia in young immunocompetent patients. The Lancet 359 753-759. Guignard, B., Entenza, J.M., Moreillon, P. 2005. (3-Lactams against methicillin-resistant Staphylococcus aureus. Current Opinion in Pharmacology 5 479-489. Hamilton-Miller, J.M.T. and Shah, S. 1999. Disorganization of cell division of methicillin-resistant Staphylococcus aureus by a component of tea {Camellia sinensis): a study by electron microscopy. FEMS Microbiology Letters 176 463-469. -240 Haraguchi, H., Ishikawa, H., Sakai, S., Ying, B.P., Kubo, I. 1996a. Inhibition of lipid peroxidation by a diterpenoid from Podocarpus nagi. Experientia 52 573-576. Haraguchi, H., Oike, S., Muroi, H., Kubo, I. 1996b. Mode of antibacterial action of totarol, a diterpene from Podocarpus nagi. Planta Medica 62 122-125. Haraguchi, I., Ishikawa, H., Kubo, I. 1997. Antioxidative action of diterpenoids from Podocarpus nagi. Planta Medica 63 213-215. Harrison, L.J. Asakawa, Y. 1987. 18-Oxoferruginol from the leaf of Torreya nucifera. Phytochemistry 26 1211-1212. Hasegawa, S. and Hirose, Y. 1981. Terpeonids from the seed of Thujopsis dolabrata. Phytochemistry 20 508-510. Hashimoto, T., Kumazawa, S., Najno, F., Hara, Y., Nakayama, T. 1999. Interaction of tea catechins with lipid bilayers investigated with liposome systems. Bioscience, Biotechnology and Biochemistry 63(12) 2252-2255. He, K., Zeng, L., Shi, G., Zhao, G., Kolzlowski, J.F., McLaughlin, J.L. 1997. Bioactive compounds from Taiwania cryptomerioides. Journal of Natural Products 60 38-40. Heinrich, M., Bames, J., Gibbons, S., Williamson, E.M. 2004. Fundamentals of Pharmacognosy and Phytotherapy. Churchill Livingstone, Edinburgh, pp 60-105. -241 - Heukelbach, J. and Feldmeier, H. 2004. Ectoparasites - the underestimated realm. The Lancet 363 889-891. Hileman, B. 2005. Antiseptic soaps under scrutiny. Chemical & Engineering News 83(43) 14. Hill, N., Moor, G., Cameron, M.M. Butlin, A., Preston, S., Williamson, M.S., Bass, C. 2005. Single blind, randomised, comparative study of the Bug Buster kit and over the counter pediculicide treatments against head lice in the United Kingdom. British Medical Journal 331 384-387. Hiramatsu, K., Cui, L., Kuroda, M., Ito, T. 2001. The emergence and evolution of methicillin-resistant Staphylococcus aureus. Trends in Microbiology 9(10) 486-493. Hirose, Y., Hasegawa, S., Ozaki, N. 1983. Three new terpenoid quinone methides from the seeds of Chamaecyparis obtusa. Tetrahedron Letters 24 1535-1538. Hostettmann, K., Marston, A., Hostettmann, M. 1998. Preparative Chromatography Techniques: Applications in Natural Product Isolation. 2"^ Ed. Springer-Verlag, Berlin, pp 3-14. Houghton, P.J., and Raman, A. 1998. Laboratory Handbook for the Fractionation of Natural Extracts. Chapman & Hall, London, pp 22-53. -242- Hsu, K.C., Fang, J.M., Cheng Y.S. 1995. Diterpenes from pericarps of Chamaecyparis formosensis. Journal of Natural Products 58(10) 1592-1595. http://www.dh.gov.uk/PublicationsAndStatistics/Publications/PublicationsStatistics/Publ icationsStatisticsArticle/fs/en?CONTENT_ID=4085951 &chk=HBt2QD http://www.idsociety.org/Content/ContentGroups/Public_Statements_and_Policiesl/Stat ements/Bad_Bugs,_No_Drugs_As_Antibiotic_Discovery_Stagnates_A_Public_Health_ Crisis_Brews.htm http://www.nao.org.uk/pn/03-04/0304876.htm Hunter, J.A. and Barker, S.C. 2003. Susceptibility of head lice {Pediculus humanus capitis) to pediculicides in Australia. Parasitology Research 90 476-478. Imaizumi, Y., Sakamoto, K., Yamada, A., Hotta, A., Ohya, S., Muraki, K., Uchiyama, M., Ohwada, T. 2002. Molecular basis of pimarane compounds as novel activators of large-conductance Ca^^-activated channel alpha-subunit. Molecular Pharmacology 62 836-846. Iwamoto, M., Minami, T., Tokuda, H., Ohtsu, H., Tanaka, R. 2003. Potential antitumor promoting diterpenoids from the stem bark of Thuja standishii. Planta Medica 69 69-72. -243 Jacobson, M., Redfem, R.E., Mills, G.D. 1975. Naturally occurring insect growth regulators part 2 screening of insect and plant extracts as insect juvenile hormone mimics. Lloydia. 38(6) 455-472. Jang, Y.S., Lee, C.H., Kim, M.K., Kim, J.H., Lee, S.H., Lee, H.S. 2005. Acaricidal activity of active constituents isolated in Chamaecyparis obtusa leaves against Dermatophagoides spp. Journal o f Agricultural and Food Chemistry 53(6) 1934-1937. Jevons, M.P. 1961. “Celbenin”-resistant Staphylococci. British Medical Journal 1 124- 125. Johnson, A.?., Aucken, H.M., Cavendish, S., Ganner, M., Wale, M.C.J., Warner, M., Livermore, D.M., Cookson, B.D. 2001. Dominance of EMRSA-15 and -16 among MRSA causing nosocomial bacteraemia in the UK: analysis of isolates from the European Antimicrobial Resistance Surveillance System (EARSS). Journal o f Antimicrobial Chemotherapy 143-144. Johnston, W.H., Karchesy, J.J., Constantine, G.H., Craig, A.M. 2001. Antimicrobial activity of some pacific northwest woods against anaerobic bacteria and yeast. Phytotherapy Research 15 586-588. Kaatz, G.W., Seo, S.M., O’Brien, L., Wahiduzzaman, M., Foster, T.J. 2000. Evidence for the existence of a multidrug efflux transporter distinct from Nor A in Staphylococcus aureus. Antimicrobial Agents and Chemotherapy 44(5) 1404-1406. -244- Kaatz, G.W., Seo, S.M., Ruble, C.A. 1993. Efflux-mediated fluoroquinolone resistance in Staphylococcus aureus. Antimicrobial Agents and Chemotherapy 37(5) 1086-1094. Kabouche, A., Boutaghane, N., Kabouche, Z., Seguin, E., Tillequin, P., Benlabed, K. 2005. Components and antibacterial activity of the roots of Salvia jaminiana. Fitoterapia 76 450-452. Kahne, D., Leimkuhler, C., Lu, W., Walsh, C. 2005. Glycopeptide and lipoglycopeptide antibiotics. Chemical Reviews 105(2) 425-448. Karawya, M.S., Soliman, P.M., Aboutabl, E.A., el-Kersh, T.A. 1986. Volatile oil of Chamaecyparis lawsoniana Pari. Egyptian Journal of Pharmaceutical Sciences 27 1-4. Karlberg, A.T. 1989. Pure abietic acid is not allergenic. Contact Dermatitis 21 282-284. Karlberg. A.T., Bergstedt, E., Boman, A., Bohlinder, K., Liden, C., Nilsson, J.L.G., Wahlberg, I.E. 1985. Is abietic acid the allergenic component of colophony? Contact Dermatitis 13 209-215. Katz, L. and Ashley, G.W. 2005. Translation and protein synthesis: macrolides. Chemical Reviews 105 499-527. 245 Kelecom, A. 1984. An abietane diterpene from the labiate Coleus barbatus. Phytochemistry 2'^ 1677-1679. Kemp, W. 1978. Organic Spectroscopy. The MacMillan Press Ltd., London, pp 76-152. Kmietowicz, Z. 2003. Information for patients: removal of lice and eggs by combing. British MedicalJournal 326 1258. Kobayashi, K., Nishino, C., Fukushima, M., Shiobara, Y., Kodama, M. 1988. Antibacterial activity of pisiferic acid and its derivatives against Gram-negative and Gram-positive bacteria. Agricultural and Biological Chemistry 52(1) 77-83. Kobayashi, K., Nishino, C., Tomita, H. Fukushima, M. 1987. Antifungal activity of pisiferic acid derivatives against the rice blast fungus. Phytochemistry 26 3175-3179. Kopper, B.J., Illman, B.L. Kersten, P.J., Klepzig, K.D., Raffa, K.F. 2005. Effects of diterpene acids on components of a conifer bark beetle-fungal interaction: tolerance by Ips pini and sensitivity by its associate Ophiostoma ips. Environmental Entomology 34(2) 486-493. Koyama, S., Yamaguchi, Y., Tanaka, S., Motoyoshiya. 1997. A new substance (Yoshixol) with an interesting antibiotic mechanism from wood oil of Japanese traditional tree (Kiso-Hinoki), Chamaecyparis obtusa. General Pharmacology 28(5) 797-804. -246- Kubo, L, Muroi, H., Kubo, A. 1995. Structural functions of antimicrobial long-chain alcohols and phenols. Bioorganic and Medicinal Chemistry 3(7) 873-880. Kubo, I., Muroi, H., Himejima, M. 1992. Antibacterial activity of totarol and its potentiation. Journal of Natural Products 55(10) 1436-1440. Kuo, Y.H. and Chan, Y.F. 2005. Six new components from the leaves of Chamaecyparis obtusa var. formosana. Abstract (P288), Society o f Medicinal Plant Research Conference, Florence, Italy, 21-25 August 2005. Kuo, Y.H., Chen, C.H., Lin Y.L. 2002. New lignans from the heartwood of Chamaecyparis obtusa var. formosana. Chemical and Pharmaceutical Bulletin 50(7) 978-980. Lahlou, M. 2004. Methods to study the phytochemistry and bioactivity of essential oils. Phytotherapy Research 18 435-448. Langenheim, J.H. 2003. Plant Resins: Chemistry, Evolution, Ecology and Ethnobotany. Timber Press Inc. Portland, Oregon, USA. Lechner D, Stavri M, Oluwatuyi M, Pereda-Miranda R, Gibbons S. 2004. The anti- staphylococcal activity o ïAngelica dahurica (Bai Zhi). Phytochemistry 65 331-335. -247- Leeb, M. 2004. A shot in the arm. Nature 431 892-893 Leo, N.P., Hughes, J.M., Yang, X., Poudel, S.K.S., Brogdon, W.G., Barker, S.C. 2005. The head and body lice of humans are genetically distinct: evidence from double infestations. Heredity 95(1) 34-40. Leung, A.Y. and Foster, S. 1996. Encyclopedia of common natural ingredients used in food, drugs and cosmetics. 2^^ Ed. John Wiley & Sons, Inc., New York, pp 502-504. Levy, C.W., Roujeinikova, A., Sedelnikova, S., Baker, P.J., Stuitje, A., Slabas, A.R., Rice, D.W., Rafferty, J.B. 1999. Molecular basis of Triclosan activity. Nature 398 383- 384. Levy, S.B. 2005. Antibiotic resistance - the problem intensifies. Advanced Drug Delivery Reviews SI 1446-1450. Li, X.Z. and Nikaido, H. 2004. Efflux-mediated drug resistance in bacteria. Drugs 64(2) 159-204. Lien, E.J., Hansch, C., Anderson, S.M. 1968. Structure-activity correlations for antibacterial agents on Gram-positive and Gram-negative cells. Journal of Medicinal Chemistry 11 430-441. -248- Lin, T.C., Fang, J.M., Cheng, Y.S. 1999. Terpenes and lignans from leaves of Chamaecyparis formosensis. Phytochemistry 51 793-801. Lin, T.S., Chang, T.T. 1999. Termite and decay resistance of two imported Canadian and three domestic woods. Taiwan Journal of Forest Science 14(2) 235-239. Lindsay, J.A and Holden, M.T.G. 2004. Staphylococcus aureus: superbug, super genome? Trends in Microbiology 12(8) 378-385. Little, D.P., Schwarzbach, A.E., Adams, R.P., Hsieh, C.F. 2004. The circumscription and phylogenetic relationships of Callitropsis and the newly described genus Xanthocyparis (Cupressaceae). American Journal of Botany 91(11) 1872-1881. Liu, X.M. Wei, J.P., Tan, F.S., Zhou, S.M. Wurthwein, G., Rohdewald, P. 2004. Antidiabetic effect of Pycnogenol® French maritime pine bark extract in patients with diabetes type II. Life Sciences 75(21) 2505-2513. Livermore, D.M. and Williams, J.D. 1996. P-Lactams: mode of action and mechanisms of bacterial resistance. Chapter 12. Antibiotics in Laboratory Medicine., ed. V. Lorian, 4* ed. Williams & Wilkins Co., Baltimore, USA. Luo, H.W., Hu, X.J., Wang, N., Ji, J. 1988. Platelet aggregation inhibitors from Salvia militiorrhiza Bunge. Acta Pharmacologica Sinica 23(11) 830-834. 249- Mabey, R. 1996. Flora Britannica. Sinclair Stevenson, London, pp 24. Macchioni, F., Cioni, P.L., Flamini, G., Morelli, I., Maccioni, S., Ansaldi, M. 2003. Chemical composition of essential oils from needles, branches and cones of Pinus pinea, P. halepensis, P. pinaster and P. nigra from central Italy. Flavour and Fragrance Journal 18 139-143. Mallorqui-Femandez, G., Mairero, A., Garcia-Piquè, S., Garcia-Castellanos, R., Gomis- Rüth, F.X. 2004. Staphylococcal methicillin resistance: fine focus on folds and functions. FEMS Microbiology Letters 235 1-8. Mangoni, L. and Caputo, R. 1967. Sempervirol, a novel type of diterpene phenol. Tetrahedron Letters 8 673-675. Markham, P.N., Westhaus, E., Klyachko, K., Johnson, M.E. and Neyfakh, A.A. 1999. Multiple novel inhibitors of the Nor A multidrug transporter of Staphylococcus aureus. Antimicrobial Agents and Chemotherapy 43 2404-2408. Marshall, C., Wesselingh, S., McDonald, M., Spelman, D. 2004. Control of endemic MRSA - what is the evidence? A personal view. Journal o f Hospital Infection 56 253- 268. Marshall, N.J. and Piddock, L.J.V. 1997. Antibacterial efflux systems. Microbiologia 13 285-300. -250- McDaniel C.A. 1989. Major termiticidal components of heartwood of Port-Orford cedar, Chamaecyparis lawsoniana (A. Murr.) Pari. Material und Organismen 24(1) 1-16. McKeegan, K.S., Borges-Walmsley, M.I., Walmsley, A.R. 2004. Structural understanding of efflux-mediated drug resistance: potential routes to efflux inhibition. Current Opinion in Pharmacology 4 479-485. Meka, V.G. and Gold, H.S. Antimicrobial resistance to linezolid. 2004. Clinical Infectious Diseases 39 1010-1015, Micol, V., Mateo, C.R., Shapiro, S., Aranda, F.J., Villalain,J. 2001. Effects of (+)- totarol, a diterpenoid antibacterial agent, on phospholipid model membranes. Biochimica et Biophysica Acta 1511 281-290. Morsy, T.A., Abou el Ela, R.G.H., Nasser, M.M.I., Khalaf, S.A.A., Mazyad, S.A.M. 2000. Evaluation of the in vitro pediculicidal action of four known insecticides and three medicinal plant extracts. Journal of the Egyptian Society o f Parasitology 30(3) 699-708. Mosmann, T. 1983. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. Journal of Immunological Methods 65 55-63. -251 - Mossa, J.S., El-Feraly, F.S. Muhammad, I. 2004. Antimycobacterial constitutents from Juniperus procera, Ferula communis and Plumbago zeylanica and their in vitro synergistic activity with isonicotinic acid hydrazide. Phytotherapy Research 18 934-937. Muhammad, I., Mossa, J.S., A1 Yahya, M.A., Ramadan, A.F., El-Feraly, F.S. 1995. Further antibacterial diterpenes from the bark and leaves of Juniperus procera Hochstex Endl. Phytotherapy Research 9 584-588. Muhammad, I., Mossa, J.S., El-Feraly, F.S. 1992. Antibacterial diterpenes from the leaves and seeds oî Juniperus excelsa M.Bieb. Phytotherapy Research 6 261-264. Muroi, H. and Kubo, I. 1996. Antibacterial activity of anacardic acid and totarol, alone and in combination with methicillin against methicillin-resistant Staphylococcus aureus. Journal of Applied Bacteriology 80 387-394. Nakanishi, T., Miyasaka, H., Nasu, M., Hasimoto, H., Yoneda, K. 1983. Production of cryptotanshinone and ferruginol in cultured cells of Salvia miltiorrhiza. Phytochemistry 22 721-722. NCCLS - National Committee for Clinical Laboratory Standards 1998. M100-S8 Performance Standards for Antimicrobial Susceptibility Testing. Vol. 18 No.l pp 56-59. 252- Nicolaou, K.C., Yang, Z., Liu, J.J., Ueno, H., Nantermet, P.O., Guy, R.K., Claiborne, C.F., Renaud J., Couladouros, E.A., Paulvannan, K., Sorensen, E J. 1994. Total synthesis of taxol. Nature 367 630-634. Nicolson, K., Evans, G., O’Toole, P.W. 1999. Potentiation of methicillin activity against methicillin-resistant Staphylococcus aureus by diterpenes. FEMS Microbiology Letters 179 233-239. Nikinmaa, M., Wickstrom, C., Lilius, H., Isomaa, B., Rabergh, C. 1999. Effects of resin acids on hepatocyte pH in rainbow trout (Oncorhynchus mykiss). Environmental Toxicology and Chemistry 18 993-997. Nishida,T., Wahlberg, I., Enzell, C.R. 1977. Carbon-13 nuclear magnetic resonance spectra of some aromatic diterpenoids. Organic Magnetic Resonance 9(4) 203-209. Oluwatuyi, M., Kaatz, G.W., Gibbons, S. 2004. Antibacterial and resistance modifying activity of Rosmarinus officinalis. Phytochemistry 65 3249-3254. Overbye, K.M. and Barrett, J.F. 2005. Antibiotics: where did we go wrong? Drug Discovery Today 10(1) 45-52 Ozaki, N., Hasegawa, S., Hirose, Y. 1983. Terpenoids from the seed of Chamaecyparis obtusa. Phytochemistry 22 1771-1773. -253 Panella, N.A., Dolan, M.C., Karchesy, J., Xiong, Y.P., Peralta-Cruz, J., Khasawneh, M, Montenieri, J.A., Maupin, G.O. 2005. Use of novel compounds for pest control: insecticidal and acaricidal activity of essential oil components from heartwood of Alaska yellow cedar. Journal of Medical Entomology 42(3) 352-358. Panella, N.A., Karchesy, J., Maupin, G.O., Malan, J.C., Piesman, J. 1997. Susceptibility of immature Ixodes scapularis (Acari: Ixodidae) to plant-derived acaricides. Journal o f Medical Entomology 34 (3) 340-345. Paulsen, I.T. 2003. Multidrug efflux pumps and resistance: regulation and evolution. Current Opinion in Microbiology 6 446.451. Pelletier, S.W., Chokshi, H.P., Desai, H.K. 1986. Separation of diterpenoid alkaloid mixtures using vacuum liquid chromatography. Journal of Natural Products 49(5) 892- 900. Pettit, G.R., Rui, T., Northen, J.S., Herald, D.L., Chapuis, J.C., Pettit, R.K. 2004. Antineoplastic agents. 529. Isolation and structure of nootkastatins 1 and 2 from the Alaskan Yellow Cedar Chamaecyparis nootkatensis. Journal of Natural Products 67 1476-1482. Percival, S.L., Bowler, P.G., Russell, D. 2005. Bacterial resistance to silver in wound care. Journal of Hospitallnfection 60 1-7. -254- Phillips, M.A. and Croteau, R.B. 1999. Resin-based defenses in conifers. Trends in Plant Science 4{5) 184-190. Politi, M., Braca, A., De Tommasi, N., Morelli, I, Manunta, A., Battinelli, L., Mazzanti, G. 2003. Antimicrobial diterpenes from the seeds of Cephalotaxus harringtonia var. drupacea. Planta Medica 69 468-470. Poole, K. 2001. Overcoming antimicrobial resistance by targeting resistance mechanisms. Journal of Pharmacy and Pharmacology 53 283-294. Priestley C. PhD thesis. School of Pharmacy, University of London, 2002. Projan, S. 2003. Why is big pharma getting out of antibacterial drug discovery? Current Opinion in Microbiology 6 427-430. Qian, P., Wei, D., Zhang, Q., Yang, S. 2005. Modulation of P-glycoprotein function and reversal of multidrug resistance by (-)-epigallocatechin gallate in human cancer cells. Biomedicine & Pharmacotherapy. 59 64-69. Rabergh, C M., Lilius, H., Eriksson, I.E., Isomaa, B. 1999. The resin acids dehydroabietic acid and isopimaric acid release calcium from intracellular stores in rainbow trout hepatocytes. Aquatic Toxicology 46 55-65. 255 - Raffa, K.F., Havill, N.P., Nordheim, E.V. 2002. How many choices can your test animal compare effectively? Evaluating a critical assumption of behavioural preference tests. Oecologia 133 422-429. Richardson, J.F. and Reith, S. 1993. Characterisation of a strain of methicillin-resistant Staphylococcus aureus (EMRSA-15) by conventional and molecular methods. Journal o f Hospital Infection 25 45-52. Ritch-Krc, E.M., Turner, N.J., Towers, G.H.N. 1996. Carrier herbal medicine: an evaluation of the antimicrobial and anticancer activity in some frequently used remedies. Journal of Ethnopharmacology 151-156.52 Robards, K., Haddad, P R., Jackson, P.E. 1994. Principles and Practice of Modem Chromatographic Methods. Academic Press Ltd, London. 76-177. Roberts, R.J. and Burgess, I.F. 2005. New head-lice treatments: hope or hype? The Lancet 365 8-10. Rohdewald, P. 2002. A review of the French maritime pine bark extract (Pycnogenol), a herbal medication with a diverse clinical pharmacology. International Journal of Clinical Pharmacological Therapy 4^ 158-168. 256- Ross, J.L, Farrell, A.M., Eady, E.A., Cove, J.H., Cunliffe, W J. 1989. Characterisation and molecular cloning of the novel macrolide streptogramin B resistance determinant from Staphylococcus epidermidis. Journal of Antimicrobial Chemotherapy 24 851-862. Sakagami, H., Kawazoe, Y., Komatsu, N., Simpson, A., Nonoyama, M., Konno, K., Yoshida, T., Kuroiwa, Y., Tanuma, S.I. 1991. Antitumor, antiviral and immunopotentiating activities of pine cone extracts: potential medicinal efficacy of natural and synthetic lignin-related materials. Anticancer Research 11 881-888. San Feliciano, A., Medarde, M., Gordaliza, M., Lucas, M.J. 1995. Structure elucidation of germacrane alcohols from Juniperus communis subsp. Hemisphaerica. Journal of Natural Products 58(7) 1059-1064. San Feliciano, A., Gordaliza, M., Salinero, M.A., Miguel del Corral, J.M. 1993. Abietane acids: sources, biological activities and therapeutic uses. Planta Medica 59 485-490. San Feliciano, A. and Lopez, J.L. 1991. Recent chemistry of conifer terpenoids. Ecological Chemistry and Biochemistry of Plant Terpenoids ed. J.B. Harbome and F.A. Tomas-Barberan. Oxford University Press, England, pp 1-27. Scheffrahn, R.H., Hsu, R.C., Su, N.Y., Huffman, J.B., Midland, S.L., Sims, J.J. 1988. Allelochemical resistance of bald cypress Taxodium distichum heartwood to the 257- subterranean termite Coptotermes formosanus. Journal of Chemical Ecology 14(3) 765- 776, Schiff, P.B., Fant, J., Horwitz, S.B. 1979. Promotion of microtubule assembly in vitro by taxol. Nature 277 665-667. Senior, K. 2005. New antibiotic approved for MRSA. Lancet Infectious Diseases 5 474. Shapiro, S. and Guggenheim, B. 1998. Inhibition of oral bacteria by phenolic compounds. Part 1. QSAR analysis using molecular connectivity. Quantitative Structure Activity Relationships. 17 327-337. Sharp, H., Latif, Z., Bartholomew, B., Bright, C., Jones, C.D., Sarker, S. 2001. Totarol, totaradiol and ferruginol: three diterpenes from Thuja plicata (Cupressaceae). Biochemical Systematics and Ecology 29 215-217. Shimatani, Y., Sakamoto, O., Hayashi, M. 1989. Bactericidal compositions containing sempervirol. Patent application, Shiseido, Co., Ltd., Japan. Shibata, H., Kondo, K., Katsuyama, R., Kawazoe, K., Sata, Y., Murakami, K., Takaishi, Y., Arakaki, N., Higuti T. 2005. Alkyl gallates, intensif!ers of p-lactam susceptibility in methicillin-resistant Staphylococcus aureus. Antimicrobial Agents and Chemotherapy 49(2) 549-555. -258 Shiota, S., Shimizu, M., Mizushima, T., Ito, H., Hatano, T., Yoshida, T., Tsuchiya, T. 1999. Marked reduction in the minimum inhibitory concentration (MIC) of beta-lactams in methicillin-resistant Staphylococcus aureus produced by epicatechin gallate, an ingredient of green tea {Camellia sinensis). Biological & Pharmaceutical Bulletin 22 1388-1390. Silliman, K., Parry, J., Kirk, L.L. Prior, R.L. 2003. Pycnogenol does not impact the antioxidant or vitamin C status of healthy young adults. Journal of the American Dietetic Association 103 67-72. Smith, E.C.J., Kaatz, G.W., Seo, S.M., Williamson, E.M., Gibbons, S. 2006. Totarol inhibits multidrug resistance in efduxing strains of S. aureus, (in preparation for submission to Planta Medica). Smith, E., Williamson, E., Zloh, M., Gibbons, S. 2005. Isopimaric acid from Pinus nigra shows activity against multidrug-resistant and EMRSA strains of Staphylococcus aureus. Phytotherapy Research 19 538-542. Soderberg, T.A., Gref, R., Holm, S., Elmros, T., Hallmans, G. 1990. Antibacterial activity of rosin and resin acids in vitro. Scandinavian Journal o f Reconstructive Hand Surgery 24 199-205. 259- Son, K.H., Oh, H.M., Choi, S.K., Han, D.C., Kwon B.M. 2005. Anti-tumor abietane diterpenes from the cones of Sequoia sempervirens. Bioorganic & Medicinal Chemistry Letters 15 2019-2021. Song, Z., Liang, Z., Liu, X. 1995. Chemical characteristics of oleoresins from Chinese pine species. Biochemical Systematics and Ecology 23(5) 517-522. Spellberg, B., Powers, J.H., Brass, E.P., Miller, L.G., Edwards, I.E. 2004. Trends in antimicrobial drug development: implications for the future. Clinical Infectious Diseases 38 1270-1286. Stead, P. 1998. Methods in Biotechnology, Vol. 4: Natural Products Isolation., Ed. Canned, R.J.P. Humana Press Inc., Totowa, N.J. pp 165-208. Stermitz, F.R., Lorenz, P., Tawara, J.N., Zenewicz, L.A., Lewis, K. 2000. Synergy in a medicinal plant: antimicrobial action of berberine potentiated by 5'- methoxyhydnocarpin, a multidrug pump inhibitor. Proceedings of the National Academy o f Sciences o f the United States o f America 97 1433-1437. Still, W.C., Kahn, M., Mitra, A. 1978. Rapid chromatographic technique for preparative separations with moderate resolution. Journal o f Organic Chemistry 43(14) 2923-2925. Su, W.C., Fang, J.M., Cheng, Y.S. 1994. Labdanes from Cryptomeria japonica. Phytochemistry 37 1109-1114. -260- Suzuki, E., Hiramatsu, K., Yokota, T. 1992. Survey of methicillin-resistant clinical strains of coagulase-negative staphylococci for mecA gene distribution. Antimicrobial Agents and Chemotherapy 36(2) 429-434. Takahashi, K., Nagahama, S., Nakashima, T., Suenaga, H. 2003. Chemotaxonomy on the leaf constituents of Thujopsis dolabrata Sieb. et Zucc. - analysis of acidic extracts. Biochemical Systematics and Ecology 31(7) 723-738. Tan, N., Kaloga, M., Radtke, O.A., Kiderlen, A.F., Ôksüz, S., Ulubelen, A., Kolodziej, H. 2002. Abietane diterpenoids and triterpenoic acids from Salvia cilicica and their antileishmanial activities. Phytochemistry 61 881-884. Tchuendem, M.H.K., Mbah, J.A., Tsopmo, A., Ayafor, J.F., Sterner, O., Okunjic, C.C., Iwu, M.M., Schuster, B.M. 1999. Anti-plasmodial sesquiterpenoids from the African Reneilmia cincinnata. Phytochemistry 52 1095-1099. Tegos G, Stermitz FR, Lomovskaya O, Lewis K. 2002. Multi-drug pump inhibitors uncover remarkable activity of plant antimicrobials. Antimicrobial Agents and Chemotherapy 46 3133-3141. Tezuka, Y., Kasimu, R., Basnet, P., Namba, T., Radota, S. 1997. Aldose reductase inhibitory constituents of the root of Salvia miltiorhiza Bunge. Chemical & Pharmaceutical Bulletin 45 1306-1311. -261 - Torras, M.A.C., Faura, C.A., Schoniau, F., Rohdewald, P. 2005. Antimicrobial activity of Pycnogenol®. Phytotherapy Research 19 647-648. Truong-Bolduc, Q.C., Dunman, P.M., Strahilevitz, J., Projan, S.J. Hooper, D.C. 2005. MgrA is a multiple regulator of two new efflux pumps in Staphylococcus aureus. Journal of Bacteriology 187 2395-2405. Turner, N.J. 1988. Ethnobotany of coniferous trees in Thompson and Lillooet Interior Salish of British Columbia. Economic Botany 42(2) 177-194. Ucar, G. and Fengel D. 1995. Variation in composition of extractives from wood of Pinus nigra varieties. Phytochemistry 38 877-880. Ulubelen, A., Ôksüz, S., Kolak, U., Birman, H., Voelter, W. 2000. Cardioactive terpenoids and a new rearranged diterpene from Salvia syriaca. Planta Medica 66 627- 629. Ulubelen, A., Topçu, G. Chai, H.B., Pezzuto, J.M. 1999a. Cytotoxic activity of diterpenoids isolated from Salvia hypargeia. Pharmaceutical Biology 37(2) 148-151. Ulubelen, A., Ôksüz, S., Kolak, U., Tan, N., Bozok-Johansson C., Çelik, C., Kohlbau, H.J., Voelter, W. 1999b. Diterpenoids from the roots of Salvia bracteata. Phytochemistry S2 1455-1459. -262- Ulubelen, A., Topçu, G., Johansson, C.B. 1997. Norditerpenoids and diterpenoids from Salvia multicaulis with antituberculosis activity. Journal of Natural Products 60 1275- 1280. Ulusu, N.N., Ercil, D., Sakar, M.K., Tezcan, E.F. 2002. Abietic acid inhibits lipoxygenase activity. Phytotherapy Research 16 88-90. Villalain, J. 1997. Location of the toxic molecule abietic acid in model membranes by MAS-NMR. Biochimica et Biophysica Acta 1328 281-289. Vidakovic, M. 1991. Conifers morphology and variation. Graficki Zavod Hrvatske, Zagreb, Croatia, pp 139-179; 372-377; 491-494. Walsh, C. and Wright, G. 2005. Antibiotic resistance. Chemical Reviews 105(2) 391- 393. Wenkert, E., Buckwalter, B.L., Burfitt, I.R., Gasic, M.J., Gottlieb, H.E., Hagaman, E.W., Schell, P.M., Wovkulich, P.M. 1976. Carbon-15 nuclear magnetic resonance spectroscopy o f naturally occurring substances. Topics in Carbon-15 NMR spectroscopy (Chapter 2). Ed. George C. Levy. John Wiley & Sons Inc. pp 98-100. 263 Wenkert, E. and Buckwalter, B.L. 1972. Carbon-13 nuclear magnetic resonance spectroscopy of naturally occurring substances. X. Pimaradienes. Journal o f the American Chemical Society 94(12) 4367-4369. Williamson, E.M., Okpako, D.T., Evans, F.J. 1996. Selection, Preparation and Pharmacological Evaluation of Plant Material. John Wiley & Sons Ltd., Chichester, England, pp 15-23. Woldemichael, G.M., Wachter, G., Singh, M.P., Maiese, W.M., Timmermann, B.N. 2003. Antibacterial diterpenes from Calceolaria pinifolia. Journal of Natural Products 66 242-246. Xiao, D., Kuroyanagi, M., Itani, T., Matsuura, H., Udayama, M., Murakami, M., Umehara, K., Kawahara, N. 2001. Studies on constituents from Chamaecyparis pisifera and antibacterial activity of diterpenes. Chemical and Pharmaceutical Bulletin 49(11) 1479-1481. Yamamoto, H., Asano, N., Sawano, C., Sone, T., Gasha, T., Ono, Y. 1997. Diterpenes isolated from the resin of the resinous stem canker of Japanese Cypress, Chamaecyparis obtusa. Mokuzai Gakkaishi 43 558-565. Yang, Z., Kitano, Y., Chiba, K., Shibata, N., Kurokawa, H., Doi, Y., Arakawa, Y., Tada, M. 2001. Synthesis of variously oxidized abietane diterpenes and their antibacterial activities against MRSA and VRE. Bioorganic and Medicinal Chemistry 9 347-356. -264- Yatagai, M. and Nakatani, N. 1994. Antimite, antifly, antioxidative and antibacterial activities of pisiferic acid and its congeners. Mokuzai Gakkaishi 40(12) 1355-1362. Yatagai, M. and Takahashi, T. 1979. Diterpenes of the ferruginol type from Chamaecyparis pisifera. Phytochemistry 18 176. Yatagai, M., Shirato, T., Hayashi, Y., Fukuhara, N., Takahashi, T. 1978. Pisiferic acid, a new phenolic diterpene carboxylic acid from Chamaecyparis pisifera S. et Z. Mokuzai Gakkaishi 24(4) 267-269. Ying, B.P. and Kubo, I. 1991. Complete *H and NMR assignments of totarol and its derivatives. Phytochemistry 30 1951-1955. Youssef, M.Y., Sadaka, H.A., Eissa, M.M., el-Ariny, A.F. 1995. Topical application of ivermectin for human ectoparasites. American Journal of Tropical Medicine and Hygiene. 53(6) 652-653. Zetola, N., Francis, J.S., Nuemberger, E.L., Bishai, W.R. 2005. Community-acquired methicillin-resistant Staphylococcus aureus: an emerging threat. Lancet Infectious Diseases 5 275-286. Zinkel, D.F., Magee, T.V., Walter, J. 1985. Major resin acids oi Pinus nigra needles. Phytochemistry 24 1273-1277. -265- Zloh, M., Kaatz, G.W., Gibbons, S. 2004. Inhibitors of multidrug resistance (MDR) have affinity for MDR substrates. Bioorganic Medicinal Chemistry Letters 14 881-885. 266- PUBLICATIONS Ming, Q., Gibbons, S., Stavri, M., Smith, E., Zhou, F-S. and Hu, C-Q. 2006. Antibacterial effect of Hypericum ascyron and Hypericum japonicum Against Multidrug-resistant (MDR) Staphylococcus aureus. Pharmaceutical Biology (Accepted). Smith, E., Williamson, E., Zloh, M., Gibbons, S. 2005. Isopimaric acid from Pinus nigra shows activity against multidrug-resistant and EMRSA strains of Staphylococcus aureus. Phytotherapy Research 19 538-542. Casely-Heyford, M., Kerr, N. O., Smith, E., Gibbons, S. and Searcey, M. 2005. Antitumour antibiotics with potent activity against multidrug-resistant Staphylococcus aureus: a new approach to targeting resistant bacteria. Medicinal Chemistry 1(6) 619-627. Gibbons, S., Moser, E., Hausmann, S., Stavri, M., Smith, E., Clennett, C. 2005. An anti-staphylococcal acylphloroglucinol from Hypericum foliosum. Phytochemistry 66(12) 1472-1475. -267-